Compositional Semantics An Introduction To The Syntaxsemantics Interface Pauline Jacobson download https://ebookbell.com/product/compositional-semantics-an- introduction-to-the-syntaxsemantics-interface-pauline- jacobson-4768428 Explore and download more ebooks at ebookbell.com
Here are some recommended products that we believe you will be interested in. You can click the link to download. Introduction To Semantics An Essential Guide To The Composition Of Meaning Thomas Ede Zimmermann Wolfgang Sternefeld https://ebookbell.com/product/introduction-to-semantics-an-essential- guide-to-the-composition-of-meaning-thomas-ede-zimmermann-wolfgang- sternefeld-50987382 Introduction To Semantics An Essential Guide To The Composition Of Meaning Thomas Ede Zimmermann https://ebookbell.com/product/introduction-to-semantics-an-essential- guide-to-the-composition-of-meaning-thomas-ede-zimmermann-4323978 An Introduction To Lexical Semantics A Formal Approach To Word Meaning And Its Composition First Published Eunhee Lee https://ebookbell.com/product/an-introduction-to-lexical-semantics-a- formal-approach-to-word-meaning-and-its-composition-first-published- eunhee-lee-48249722 Presupposition And Implicature In Compositional Semantics Uli Sauerland https://ebookbell.com/product/presupposition-and-implicature-in- compositional-semantics-uli-sauerland-5360714
Default Semantics Foundations Of A Compositional Theory Of Acts Of Communication Illustrated Edition K M Jaszczolt https://ebookbell.com/product/default-semantics-foundations-of-a- compositional-theory-of-acts-of-communication-illustrated-edition-k-m- jaszczolt-1394278 The German Perfect Its Semantic Composition And Its Interactions With Temporal Adverbials 1st Edition Renate Musan Auth https://ebookbell.com/product/the-german-perfect-its-semantic- composition-and-its-interactions-with-temporal-adverbials-1st-edition- renate-musan-auth-4481838 Service Composition For The Semantic Web Foreword By Schaharam Dustdar 2011th Edition Brahim Medjahed https://ebookbell.com/product/service-composition-for-the-semantic- web-foreword-by-schaharam-dustdar-2011th-edition-brahim- medjahed-55302634 Agentbased Semantic Web Service Composition Springerbriefs In Electrical And Computer Engineering 2012th Edition Kumar https://ebookbell.com/product/agentbased-semantic-web-service- composition-springerbriefs-in-electrical-and-computer- engineering-2012th-edition-kumar-55302636 Semantic Web Services And Web Process Composition First International Workshop Swswpc 2004 San Diego Ca Usa July 6 2004 Revised Selected Papers 1st Edition Jorge Cardoso https://ebookbell.com/product/semantic-web-services-and-web-process- composition-first-international-workshop-swswpc-2004-san-diego-ca-usa- july-6-2004-revised-selected-papers-1st-edition-jorge-cardoso-1549518
OXFORD TEXTBOOKS IN LINGUISTICS Compositional Semantics
OXFORD TEXTBOOKS IN LINGUISTICS PUBLISHED Compositional Semantics An Introduction to the Syntax/Semantics Interface by Pauline Jacobson The Grammar of Words An Introduction to Linguistic Morphology Third edition by Geert Booij A Practical Introduction to Phonetics Second edition by J. C. Catford Meaning in Language An Introduction to Semantics and Pragmatics Third edition by Alan Cruse Natural Language Syntax by Peter W. Culicover Principles and Parameters An Introduction to Syntactic Theory by Peter W. Culicover A Semantic Approach to English Grammar by R. M. W. Dixon Semantic Analysis A Practical Introduction by Cliff Goddard Pragmatics by Yan Huang The History of Languages An Introduction by Tore Janson Diachronic Syntax by Ian Roberts Cognitive Grammar An Introduction by John R. Taylor Linguistic Categorization Third edition by John R. Taylor IN PREPARATION The Lexicon An Introduction by Elisabetta Ježek Functional Discourse Analysis by Evelien Keizer Semantics and Pragmatics Meaning in Language and Discourse Second edition by Kasia M. Jaszczolt Translation Theory and Practice by Kirsten Malmkjaer Linguistic Typology Theory, Method, Data by Jae Jung Song
Compositional Semantics An Introduction to the Syntax/Semantics Interface Pauline Jacobson 1
3 Great Clarendon Street, Oxford, OX2 6DP, United Kingdom Oxford University Press is a department of the University of Oxford. It furthers the University’s objective of excellence in research, scholarship, and education by publishing worldwide. Oxford is a registered trade mark of Oxford University Press in the UK and in certain other countries © Pauline Jacobson 2014 The moral rights of the author have been asserted First Edition published in 2014 Impression: 1 All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, without the prior permission in writing of Oxford University Press, or as expressly permitted by law, by licence, or under terms agreed with the appropriate reprographics rights organization. Enquiries concerning reproduction outside the scope of the above should be sent to the Rights Department, Oxford University Press, at the address above You must not circulate this work in any other form and you must impose this same condition on any acquirer Published in the United States of America by Oxford University Press 198 Madison Avenue, New York, NY 10016, United States of America British Library Cataloguing in Publication Data Data available Library of Congress Control Number: 2013957565 ISBN 978–0–19–967714–6 (Hbk.) 978–0–19–967715–3 (Pbk.) Printed and bound by CPI Group (UK) Ltd, Croydon CR0 4YY Links to third party websites are provided by Oxford in good faith and for information only. Oxford disclaims any responsibility for the materials contained in any third party website referenced in this work.
To the memory of my parents, Florence and Nathan Jacobson (Florie and Jake)
Contents Sections and exercises marked with * are more advanced and/or less central (and on some occasions intended as open-ended speculation). These may be skipped with little or no consequence for later sections. Acknowledgments xv Foreword: On Using this Book xvii Part I: Foundational Concepts: Building a Fragment 1. Introduction 3 1.1. Goals 3 1.1.1. Compositional semantics and (some of) the goals of semantic theory 4 1.1.2. Direct Compositionality—and its role in this text 8 1.2. A brief note on the history of semantics within modern linguistic theory 10 1.3. The notion of a “fragment” and its use in this text 12 *1.4. An intriguing puzzle 13 1.5. Appendix: Sets and Functions 19 1.5.1. Sets, members, and subsets 19 1.5.2. Union, intersection, and complement 21 1.5.3. Ordered pairs, relations, equivalence relations, and partitions 21 1.5.4. Functions 24 2. Semantic foundations 27 2.1. Model-theoretic semantics 27 2.2. Truth conditions 28 2.3. Possible worlds 31 2.3.1. Introducing the notion 31 2.3.2. Characteristic function of a set 35
2.3.3. Notation and terminology 36 2.3.4. Talking about worlds 37 2.4. Times 41 3. Compositionality, Direct Compositionality, and the syntax/semantics interface 42 3.1. Building a fragment: First steps 43 3.2. Implicatures vs truth conditions 46 3.3. Folding in worlds (and times) 48 3.4. Negation: A first pass 49 4. Expanding the fragment: Syntactic categories and semantic types 52 4.1. Noun phrases 52 4.2. Intransitive verbs/verb phrases 58 4.3. A brief look at lexical semantics 60 4.4. Back to the compositional semantics 62 4.5. Illustrating the syntactic/semantic composition 63 5. Transitive verbs: Resolving an apparent syntax/semantics mismatch 66 5.1. First pass: Straw Man’s solution 67 5.2. Abandoning Direct Compositionality 70 5.3. Hidden S conjunction? 73 5.4. Back to the drawing board: Curry’ed functions 76 5.4.1. Recasting the meaning of transitive verbs 76 *5.4.2. Currying more generally 79 5.4.3. Solving Straw Man’s problems 80 *5.5. A note on three-place verbs 84 6. Categorial Grammar 88 6.1. Basics 89 6.2. Syntax/semantics correspondence 92 6.3. Refinements 94 *6.4. Some further observations about syntax: Case marking and agreement 96 6.5. Further generalizing the rules 98 6.6. and, or, negation, and case-marking prepositions 100 6.7. Summary: The fragment so far 104 viii CONTENTS
7. The autonomy of syntax? 107 7.1. Good syntax—bad semantics 108 7.2. Bad syntax—good semantics 111 8. Adjectives, nouns, determiners, and more 113 8.1. Predicative adjectives 113 8.2. Nouns (“common nouns”) 116 8.3. the 118 8.4. Prenominal adjectives and unary rules 120 8.4.1. An additional binary rule? 120 8.4.2. Interlude: Unary rules (“type shift” rules; morphological rules) 121 8.4.3. Back to adjectives 124 8.4.4. Are all noun modifier adjectives intersective? 126 8.5. Relational nouns 128 8.6. Summary: The fragment so far 131 9. Interlude: The semantics of variables and the lambda calculus 133 9.1. First-order predicate logic (modified) 134 9.1.1. Variables 134 9.1.2. Quantifiers 138 9.2. The lambda calculus 144 9.2.1. Lambda abstraction 144 9.2.2. Lambda conversion 147 9.3. Naming model-theoretic objects 151 Part II: Enriching the Domain 10. Returning to English: Generalized quantifiers 157 10.1. “NPs” with quantificational determiners 157 10.1.1. First pass: Individuals? 158 10.1.2. Second pass: Sets of individuals? 159 10.1.3. Third pass: English as first-order logic? Straw Man makes a cameo reappearance 160 10.2. Generalized quantifiers 162 10.2.1. The semantics 162 10.2.2. The syntax 165 10.2.3. Extending to the other quantified cases 166 CONTENTS ix
10.3. The determiners 167 10.4. Interlude: Scalar implicature 168 10.5. Generalized quantifiers and conjunction 173 10.5.1. VP conjunction 173 10.5.2. Generalized quantifier conjunction 175 *10.6. Domain restriction 178 11. Ordinary NPs and type lifting 181 11.1. Montague’s solution 182 11.2. Partee and Rooth’s solution: Lift as a unary rule 185 11.3. Generalizing the rule 186 11.4. The lowest types hypothesis 189 11.4.1. Some additional predictions 189 *11.4.2. The lowest types hypothesis: Processing vs grammar 192 11.5. Resolving some syntactic loose ends 194 11.6. Summary: The fragment so far 196 12. Generalized conjunction 198 12.1. Generalizing and and or 198 12.2. Remarks on noun conjunction 201 12.3. Summary: The fragment so far 203 Part III: Relative Clauses, Scopes, and Binding: Some Theoretical Controversies 13. Relative clauses: Sketching two accounts 207 13.1. Preliminaries 207 13.2. What do relative clauses modify? 209 13.2.1. The noun modification argument 209 *13.2.2. But—not quite so fast . . . 213 13.3. An account with movement and traces: Assignment functions in grammar 214 13.4. A Direct Compositional (and variable-free) analysis 221 13.4.1. Preliminaries: Function composition 221 13.4.2. Interlude: Coordination 224 13.4.3. Returning to relative clauses 227 13.4.3.1. The basic account 227 13.4.3.2. Syntactic loose ends 231 x CONTENTS
13.4.4. A refinement: The Geach rule 233 13.4.4.1. The rule and its application to relative clauses (and coordination) 233 *13.4.4.2. Generalized and as generalized g-sl? 238 13.5. Appendix: Reassuring the Appalled Syntactician 240 13.6. Summary: The fragment so far 242 14. Generalized quantifiers in object position: Two approaches 244 14.1. A tempting but inadequate solution 245 14.2. Quantifier (Lowering)/Raising 248 14.3. A Direct Compositional solution 255 14.4. Some comparisons 263 14.4.1. The generality of the solutions 263 14.4.2. Antecedent-Contained Ellipsis 265 14.4.2.1. The “textbook” argument 265 14.4.2.2. Antecedent-Contained Ellipsis as Transitive Verb Phrase Ellipsis 270 14.5. Generalized quantifiers in embedded sentences 272 14.6. Summary: The fragment so far 273 15. The interpretation of pronouns: Two accounts 274 15.1. Introductory remarks 274 15.1.1. Variables or not? 274 15.1.2. Pronominal binding: The empirical phenomenon 276 15.2. The LF (variable-ful) view 278 15.2.1. The basic account 278 15.2.2. A wrinkle: Weak Crossover effects 282 15.2.3. Free pronouns 285 15.3. A Direct Compositional and variable-free approach 286 15.3.1. Pronouns and “binding” 287 15.3.2. Free pronouns 293 *15.3.3. Generalizing the rules 294 15.3.4. Weak Crossover 298 15.3.5. What happens when there is more than one pronoun? 301 CONTENTS xi
15.3.5.1. Multiple pronouns which are “the same” 301 *15.3.5.2. And yet another possibility 303 15.3.5.3. More than one pronoun— different binders 304 15.4. More on Weak Crossover 305 15.4.1. Definite NPs and WCO 305 15.4.1.1. An apparent problem (for both theories) 305 15.4.1.2. Not a problem after all 306 15.4.1.3. Independent evidence: Ellipsis and the notion “sloppy identity” 307 *15.4.1.4. A more formal account of the strict reading 312 *15.4.2. A note on three-place verbs and WCO interactions 315 15.5. More on free “variables” 317 15.5.1. Free “variables” in the variable-ful account 317 15.5.2. Free “variables” in the variable-free theory 319 15.6. The i-within-i effect revisited: A solution 320 Appendices to Parts I–III: The full fragment 323 Appendix 1: The basic fragment 323 Foundational concepts 323 The lexicon 324 Word order rules 326 Unary and combinatory rules 327 *Appendix 2: Possible extensions of the fragment for three-place verbs using Wrap 329 The lexicon 329 Word order rule 330 Revision of the definition of functions corresponding to categories of the form A/B 330 Appendix 3: Informally discussed addition for VP Ellipsis 330 xii CONTENTS
Part IV: Further Topics 16. Negative Polarity Items, semantic strength, and scalar implicature revisited 333 16.1. The notion of Negative Polarity Items 334 16.1.1. any and ever 334 16.1.2. Minimizers 336 16.2. The distribution of NPIs 337 16.2.1. First pass 337 16.2.2. Second pass 339 16.3. The first breakthrough 340 16.3.1. Initial characterization of the environments 340 16.3.2. A note on many and few 347 16.4. But why? And how? 349 16.4.1. The insight from semantic strength 351 16.4.2. The second breakthrough: Kadmon and Landman on any 353 16.4.3. ever 356 16.5. Open questions and further discussion 358 16.6. Semantic strength and scalar implicatures 360 17. More binding phenomena 364 17.1. Functional questions 364 17.1.1. The phenomenon 364 17.1.2. An analysis with complex traces 366 17.1.3. A variable-free analysis 368 17.2. Binding and coordination (in variable-free semantics) 370 17.3. “Paycheck” pronouns (and “donkey” pronouns?) 371 17.3.1. Paychecks 371 17.3.2. A variable-free analysis of paycheck pronouns 374 17.3.3. Donkeys 376 17.4. “Hidden variables” 378 17.4.1. Hidden variables can always be complex 378 17.4.2. Translating into variable-free 382 18. Additional semantic dimensions: The semantics of focus 385 18.1. Computing alternatives compositionally 386 18.2. The meaning of only 390 CONTENTS xiii
18.3. Additional focus-sensitive words 392 18.4. Further compositional issues 393 18.5. Focus and binding 395 18.6. More generally: Multidimensional semantics 397 19. Intensionality and the syntax/semantics interface 398 19.1. Intensionality-sensitive words 399 19.2. Revising the fragment: Type-sensitive rules 403 References 407 Index 419 xiv CONTENTS
Acknowledgments This book began life as a series of handouts for my Formal Semantics course in 1998. The handouts grew and were revised over the years until I realized they were trying to become a book. I owe my greatest debt to the students in these classes. I never would have undertaken to write this book were it not for the enthusiasm and insight of my students who have always made teaching Formal Semantics just plain fun. I am also indebted to three scholars whose work has influenced my thinking in ways sometimes obvious and sometimes subtle. The influence of Barbara Partee will be obvious throughout this book. To David Dowty, I owe an understanding of how to think “semantically,” and my thinking about Categorial Grammar and its elegance as a theory of the syntax/semantics interface owes much to the work of Emmon Bach. The way of thinking that I learned directly or indirectly from these three semanticists has shaped in various ways the point of view taken in this book, and the way the material is put together. Oxford University Press provided me with three detailed and extraordin- arily helpful referee reports. Probably none of the referees will be satisfied with all of my decisions, but hopefully each will find ways in which their thoughtful comments have greatly improved the manuscript. I thank also Chris Kennedy both for very helpful comments on the manuscript and for “dry-running” parts of it in his courses. I owe a huge debt to Peter Klecha, who did a heroic job of going through the manuscript in detail—working through all of the exercises to make sure they were doable (indeed sometimes they weren’t), to make sure they were not presupposing material not yet introduced (indeed, sometimes they were), and to make sure the formalism was consistent (indeed, sometimes it wasn’t). I have been amazed at Peet’s ability to see just what a student new to the material could or could not be expected to do, and at his astute reading of the material which saved me from some embarrassing mistakes. I am also extremely indebted to Jackson Golden, who has done a fantastic
job in helping to put together the bibliography, and to Christopher Mescher for help in compiling the index. I thank NSF for funding on my grant BCS 0646081; although this is a textbook it also contains much of my own research which was funded under the grant. Finally, I thank Michael Rosen, for always encouraging this project, for putting up with its writing over the last two years, and for everything else. xvi ACKNOWLEDGMENTS
Foreword: On Using this Book Each instructor will no doubt find her or his own most useful path through this book (or parts of it), but some suggestions might be helpful. The book obviously contains more material than can realistically be covered in a single semester’s introductory formal semantics course. In teaching this material myself in introductory formal semantics courses (with a mix of graduate students and advanced undergraduates) I have found that I can teach most of the material through Part III plus Chapter 16 and a brief foray into one other of additional topics in Part IV, although I do not go through the material in the full detail given here. (One reason for writing such a book is so that students can get more detail on their own.) Hence, an instructor might choose to do (most of) Chapters 1–15, and one or two additional chapters in Part IV. One plan for this is to spend about three to four weeks on the material through Chapter 5, one week on Chapters 6 and 7 com- bined, and then approximately one chapter a week for the rest (Chapter 12 could take less than a week and Chapter 15 probably more). Depending on the students’ background, another reasonable semester’s goal might instead be to work in detail through Part II and then approach some of the material in Part III. The material would also probably work as the full text for a two- quarter or one-year semantics course (perhaps supplemented at the end with some readings from original sources), and the later material can be the basis of a more advanced course, supplemented with readings from the literature. I hope that the book might also be used by students already familiar with basic formal semantics but not familiar with the viewpoint of Direct Com- positionality (or not familiar with Categorial Grammar); such a student can systematically work through the material from what would be for them a new point of view. This could easily be part of a second-semester or second- year semantics course. There are some choice points in the order in which to read the material. I myself always teach Chapter 16 (on Negative Polarity Items) right after Chapter 10. This is because once students have learned about generalized quantifiers, they have the tools to approach the domain of Negative Polarity Items. I feel it is an especially satisfying reward—after working through
much technical apparatus—to apply the apparatus to a domain that is rich and so beautifully illustrates the relevance of the formal tools to the distri- bution of items that are entirely natural and conversational. So right after Chapter 10 there is a fork point: one can either go on to Chapter 11 or digress and read Chapter 16. I have set it up so that neither chapter (nor subsequent ones) presupposes the other except in occasional minor ways. Chapter 16 could also be read between Parts II and III. Moreover, the chapters in Part IV can be read in any order. The decision to include two theoretical points of view—the Direct Com- positional view and a view that instead uses Logical Form—poses some difficulties. While I would love to have written the book purely from the point of view of Direct Compositionality, I feel it would be doing students a disservice to not be able to read and appreciate the rich body of work that uses the notion of a Logical Form. My solution, then, was to illustrate these two different viewpoints side by side with respect to a key body of data (relative clauses, quantifiers in object positions, scopes, and binding). But this could be confusing for a student new to this material, so I have set it up in such a way that a student can—on a first pass—work through the relevant chapters carefully from just one of the points of view, and later go back to work through the other. In general, the material on the other point of view will need to be skimmed as it will sometimes contain empirical data of relevance to the discussion, but the technical details can be postponed. Another difficult decision centered on the material on binding. As is pointed out in the text, I give two versions of pronominal binding: a fairly standard one that makes use of Logical Form and variables, and one making no use of variables in the semantics set within the Direct Compos- itional framework. (This is based in part although not exclusively on my own work on variable-free semantics.) But this makes it appear as if Direct Compositionality is crucially tied in with a variable-free semantics. It is not, and in an earlier draft (as well as in the way I have often taught this material) I also exposit an intermediate position: one which is Direct Compositional but does make use of variables. In the end, I decided that to develop three approaches to a single domain was bound to be just too much for an introductory book. With some regret, then, I decided it was prudent to remove the discussion of the intermediate position. Still, I feel it is worth- while to be able to access this material—both for the sake of historical accuracy (it was a well-developed position) and to understand that the issue of whether or not to have variables is to some extent independent of xviii FOREWORD: ON USING THIS BOOK
the issue of Direct Compositionality. For those who are not fans of the variable-free approach, I don’t want to leave the impression that if variable- free semantics is wrong, so is Direct Compositionality: there is a very good Direct Compositional semantics of pronouns using variables. So in order to allow access to a discussion of that intermediate position, I have made it available on the textbook website, which is <http://sites.clps.brown.edu/ compositional.sem>. One further decision that some have found unusual is to delay the formal introduction of the lambda calculus until Chapter 9 which is the end of Part I. There is a reason for this. I want to stress that this is just a convenient notation for writing model-theoretic objects, and not some level of repre- sentation that is a crucial part of the theory. (For example, lambda conver- sion is not intended as a rule of grammar, but just as a way to convert one way of writing a model-theoretic object into another way of writing it.) To stress the difference between the notation and the actual meanings of expressions, I have always found it helpful to try to use ordinary prose in naming model-theoretic objects wherever possible, even when this prose leads to cumbersome descriptions like “the set of all sets that contain the dog-set as a subset.” Eventually, of course, English prose gets too cumber- some (and is not sufficiently unambiguous) and we need a better notation (hence the lambda calculus). But I find that insisting on prose whenever possible alongside with a clearer notation is helpful—not only for under- standing that notation is just that, but also to get a better intuition about fancy objects like sets of sets or functions from sets of sets to truth values. However, an instructor who prefers to make use of the lambda calculus earlier can move Chapter 9; it could have been placed after Chapter 6. For the student with no prior background in elementary set theory, I have included an appendix to Chapter 1 that contains all of the basic notions of set theory, ordered pairs, relations, and functions that are necessary for the later material. There is quite a bit of material there, and so I would recom- mend that a student with no relevant background just read the discussion of sets first, and then come back to subsequent notions as they appear in the text. It is often easiest to absorb this material when one has a reason to do so, and when it is in the context of the fuller semantic theory for which it is being pressed into service. A word about the sections and exercises marked with an asterisk. These are more advanced and any of them truly can be skipped (especially on a first pass through the material). They are intended to provide greater coverage and depth for the ambitious and curious student, and the starred FOREWORD: ON USING THIS BOOK xix
exercises can be quite challenging. Some—especially toward the end—are actually open-ended; I hope that a few might even spawn some interesting research topics for students. In any case, the starred exercises are there to be enjoyed and not to be frustrating, so I would recommend picking and choosing wisely according to a student’s level of interest and ambition. I do hope that some of these will be fun. Finally, as noted above, there is a website for this book at <http://sites.clps. brown.edu/compositional.sem>. It contains the supplementary material on binding and will be updated from time to time with other supplementary material, along with errata as they are discovered. xx FOREWORD: ON USING THIS BOOK
PART I Foundational Concepts: Building a Fragment
1 Introduction 1.1. Goals 3 1.2. A brief note on the history of semantics within modern linguistic theory 10 1.3. The notion of a “fragment” and its use in this text 12 *1.4. An intriguing puzzle 13 1.5. Appendix: Sets and Functions 19 1.1. Goals This book stems from a belief that linguistic semantics is a beautiful field, that the tools used to study formal semantics have yielded a rich body of results about fascinating and subtle data, that the field continues to produce exciting new insights at an impressive rate,1 and that there are simple and 1 Readers wishing a taste of many of the ongoing developments in formal semantics and in the syntax/semantics interface might want to look at the journals Linguistics and Philosophy (Springer), Natural Language Semantics (Springer), Journal of Semantics (Oxford University Press), and Semantics and Pragmatics (online journal, available at <http://semprag.org/>), among many other journals. Regular conferences at which cutting-edge research is presented include the annual Semantics and Linguistic Theory (SALT) conference, Sinn und Bedeutung (also annual), the biannual Amsterdam Colloquium for Language, Logic, and Informa- tion, and Semantics of Underrepresented Languages of the Americas, as well as most of the more general regular linguistics conferences. Of course, most of the work in these venues will not be accessible to a student just learning formal semantics, but it is hoped that this book will give a large part of the necessary background for following at least some of this research. In any case, a glance at the list of papers in
elegant tools to model how the syntax and semantics of a natural language work together. We begin with a very elementary “fragment” of English and proceed to expand it further and further—adding tools as needed but aiming to keep the basic machinery relatively simple. The goal of proceeding in this way is to account for a domain of data which is sufficiently rich as to show the excitement of studying formal semantics and its interaction with syntax. We note one limitation from the outset: this book concentrates entirely on the analysis of English. The project of modeling the semantics and the syntax/semantics interaction of any single language already provides such a rich set of results that one can hopefully find this limitation justified for an introductory book like this. In fact, the results that have been gleaned from a detailed modeling of one language have in recent years allowed the field to expand so as to provide a wealth of analyses of other languages.2 This book hopes to give the foundation to approach that literature. 1.1.1. Compositional semantics and (some of) the goals of semantic theory One of the most striking and fundamental properties of language—any language—is that speakers have the ability to produce and understand an unlimited number of expressions that they have never produced or heard before (indeed many of these will have never before been uttered by any- one). This simple point is stressed in just about every introductory linguistics textbook, often phrased this way: “A speaker of a language is able to recognize as well-formed an unlimited number of expressions that s/he has never heard before.” Examples that demonstrate this are easy to construct. One can, for instance, note the existence of expressions like those in (1): any of these venues can give the reader a taste of the richness of the domain of inquiry within linguistic semantics. 2 Much cross-linguistic semantic work can be found in the journals and confer- ence proceedings cited in footnote 1. An early edited volume on this is Bach et al.’s Quantification in Natural Languages. There is now also an annual conference Semantics of Underrepresented Languages of the Americas with published confer- ence proceedings. And many of the specialized conferences on individual languages and language families regularly include work on semantics. 4 1. INTRODUCTION
(1) a. the tallest linguistics major b. the tallest linguistics major who is graduating in December c. the tallest linguistics major who is graduating in December who is enrolled in formal semantics d. the tallest linguistics major who is graduating in December who is enrolled in formal semantics who took phonology last semester . . . One can keep forming longer and longer expressions like this by adding new relative clauses (each of the phrases that begin with who here is what is commonly known as a relative clause). But while this is often put in terms of a speaker’s ability to recognize that these are well-formed, that is surely only part of the story. Even more interesting (at least to a semanticist) is the fact that speakers know how to interpret these expressions. The rule system that speakers have unconsciously learned is hardly just a system to determine whether a given string of words is an expression of the language in question (here English); language would be quite useless if it were just a collection of meaningless strings. And so, in modeling what a speaker of English “knows” (in an uncon- scious sense, of course) about her/his language we want to predict how it is that s/he can understand expressions like those in (1) no matter how many relative clauses they contain. Thus speakers obviously have as part of their knowledge a finite set of basic items—call these the words and call the collection of the basic items the lexicon. (Here and for most of this text we ignore the distinction between words and morphemes.) Since the lexicon is finite, the meanings of the basic items can be learned on a case-by-case basis. But this obviously cannot be the case for the larger expressions: there has to be some systematic set of principles that speakers have that allows them to understand their meanings on the basis of the meanings of the smaller parts (ultimately the words) that make them up. This is the system which is called the compositional semantics—and one of the jobs of a theory of the seman- tics (of any language) is to model the rules and/or principles which allow speakers to understand an unlimited number of expressions. This book is primarily about just this. Let’s look a bit more at the expressions in (1). When a speaker utters any of these expressions—perhaps as part of a fuller sentence like in (2)—the act of uttering these expressions takes place in a fuller discourse context, and we understand them relative to facts about that context: (2) We need to make sure to order academic regalia which is long enough to fit the tallest linguistics major (who is graduating in December (who . . . )) 1.1. GOALS 5
The role of context will be discussed more formally at various points in the text, but the informal notion of a speech or discourse context is clear enough. So suppose we are using the expressions in (1) in a context in which it is obvious that we are concerned with the students at Brown University. Given this (or any other context), we can see that any speaker of English immediately knows some interesting facts about these expressions—facts which our model of the compositional semantics needs to account for. Take for instance (1a). It refers to some unique individual.3 The hearer may well not know who exactly that is—in fact the speaker might not either (as is clear in a context like (2)). But both parties assume that there is a particular individual (and only one) referred to by each of these expressions. And there are many other inferences that can be drawn from these. For example, we immediately know that if the individual described by (1a) is Nora, then either she’s also the individual described by (1b) or else she is not graduating in December. Moreover, if Nora is not the person picked out by (1b) then whoever that person is, s/he must be shorter than Nora. Similarly, with each successively longer phrase we either refer to the same person, or to one who is shorter. Suppose that Zorba is the person described by (1b). We know that he is shorter than Nora, and also know that if he is not the person described by (1c) then he is not enrolled in formal semantics. And whoever the (1c) person is—let’s say Otto—Otto must be shorter than Zorba. The addition of each successive relative clause either keeps the referent constant or allows shorter and shorter people to “rise to the top.” This kind of knowledge is automatic and immediate, and it is the job of a model of the compositional semantics to explicitly account for inferences like this. We won’t give a serious account of any of this at this point, but can hint at one possible account. Suppose that an expression like linguistics major refers to some set of individuals. (Readers not familiar with basic notions of set theory should consult the Appendix to this chapter.) When this set is put together with the tallest (pretend that the tallest is a single word here), the entire expression ends up referring to the tallest member of that set. Nothing 3 In reality there could conceivably be two individuals of exactly the same height. But use of the expressions in (1) does seem to assume that there is a unique referent for these. This is sometimes called a presupposition; these are rather odd expressions if the speaker knows that there are two individuals with exactly the same height (in that case the speaker might have said the two tallest linguistics majors). 6 1. INTRODUCTION
surprising so far. But what is more interesting is what happens with the addition of further relative clauses. It seems plausible that something like who is graduating in December also refers to a set (obviously, the set of December graduates). The above facts will make sense if the compositional semantics first combines the two sets (the set of linguistics majors and the set of December graduates) and intersects them to give a new set. (The intersection of two sets is all things that are in both sets; again see the Appendix.) So (1b) ends up picking out the tallest member of that set. It is now possible to demonstrate that the system correctly predicts that if the referent of (1b) is not Nora, it can only be because she is not graduating in December. For if Nora is taller than anyone in the linguistics major set (call that L) then she is taller than anyone in the intersection of L with the December graduates (call that D). After all, everyone who is in that intersection of L and D is also in L. So if Nora is not the referent of (1b) it can only be that she’s not in the intersection of D and L, and since she’s in L (by assumption) it follows that she can’t be in D. It also follows that if (1b) refers to Zorba, he must be shorter than Nora. By the definition of intersection, if Zorba is in the intersection of D and L he is in L, but we already know that Nora is taller than everyone else in L. All of this is very simple logic that we—the linguists—can work out in the form of an informal proof as above. It could also be worked more formally if one were so inclined. Pedantic though it may seem, it shows that our compositional procedure (which involves intersecting two sets) can be used to correctly model inferences that speakers of English effort- lessly make. Moreover, the appeal here is that this is perfectly general and extends no matter how many new relative clauses are added. Take (1c). The semantics set-up above extends immediately to this. The new relative clause in (1c) is who is enrolled in formal semantics. This picks out yet another set—and so this now intersects with the set that we already formed for (1b). The fact that the referent of (1c) can either be Zorba or someone shorter than Zorba follows by the same logic shown above; the reader can work out the details. And the procedure can be repeated over and over no matter how many relative clauses are introduced. 1.1. GOALS 7
1.1.2. Direct Compositionality—and its role in this text This book has a rather ambitious set of goals. On the one hand, I intend this to be a stand-alone text for anyone wishing to have an introduction to formal semantics, compositional semantics, or what is commonly known as the syntax/semantics interface. In other words, we will be asking (as in the above example) what a compositional semantics might look like: how can we model the tools available (again, of course, unconsciously) to speakers of a language that allow them to compute meanings of larger expressions from the meanings of the smaller ones that make them up. What are the formal ways in which meanings combine? And what are the types of objects that we need in order to model that? (For example, the discussion above shows that some simple tools of set theory can be useful.) But while most semanticists agree that (in general) the meaning of a larger expression is built in some systematic way from the meanings of the parts that make it up, just exactly how the syntactic system of a language and the compositional semantics work together is a matter of considerable controversy, and is one of the central questions addressed in this book. And so this book takes one particular point of view on this: the point of view known as Direct Compositionality. This view was explored perhaps most notably in Montague (1970) and was either generally accepted or at least taken as a serious desideratum in much of the work in linguistic formal semantics throughout the 1970s and 1980s (particularly work in what was then known as the Montague Grammar program). It was also taken as the foundation for semantics in syntactic theories such as Generalized Phrase Structure Grammar (Gazdar, Klein, Pullum, and Sag 1985), and is assumed in a large body within current grammatical theories that go under the rubric of Categorial Grammar, Type-Logical Grammar, and other related theories. 1.1. Because this example is just meant to illustrate the notion of a compositional semantics, we have made some assumptions about the order in which the semantics put things together without justifying them. Suppose that rather than the way it was set up here, the meanings of the two relative clauses (1c) first combined, and then that combined with linguistics major. Would that make any difference to the basic semantic compositional picture that we have set up here? Would the procedure extend correctly to (1d)? 8 1. INTRODUCTION
To elucidate, a fairly uncontroversial claim is that the grammar of any natural language is a system of rules (or principles, if one prefers) that define the set of well-formed expressions of the language (i.e., the syntax) and a set of rules (or principles) pairing these with meanings (i.e., the semantics). The hypothesis of Direct Compositionality is a simple one: the two systems work in tandem. Each expression that is proven well-formed in the syntax is assigned a meaning by the semantics, and the syntactic rules or principles which prove an expression as well-formed are paired with the semantics which assign the expression a meaning. (An interesting consequence of this view is that every well-formed syntactic expression does have a meaning.4 ) It is not only the case that every well-formed sentence has a meaning, but also each local expression (“constituent”) within the sentence that the syntax defines as well-formed has a meaning. Of course putting it this way is arguably not much more than just a slogan: the empirical content of this depends in part on just how the syntax works and what one takes to be a meaning. This will be filled in as we proceed. It might also seem at first glance that the hypothesis of Direct Compositionality is a fairly trivial one. But in fact it is not always immediately obvious how to give a Direct Compositional analysis. Even the example in 1.1.1 is a case in point. If the syntax and semantics work together, then the analysis given above leads to the conclusion that in the syntax a relative clause like who is graduating in December combines with linguistics major rather than with the tallest lin- guistics major. But this very question regarding the syntax of relative clauses has been debated in the literature since the 1960s, and many researchers have claimed that the syntactic constituent structure of the tallest linguistics major who is graduating next year is not the structure that was used above for the semantic analysis. We will actually revisit this particular question in later chapters (see, e.g., section 13.2). So one of the goals of this book will be to see what it takes to give Direct Compositional analyses of a variety of constructions. While the material in this book is generally exposited from the Direct Compositional point of view (along with discussion of the challenges to this hypothesis), the book is also intended to be a perfectly reasonable 4 Of course one of the earliest arguments in Generative Grammar for divorcing the syntax from the semantics (and thus a putative argument against Direct Com- positionality) is based on the claim that there are well-formed expressions that don’t have any meaning (Chomsky 1957). This is addressed in section 7.1. 1.1. GOALS 9
stand-alone textbook for any formal semantics course. Thus it is suitable for any linguistics student or linguist wanting a ground-up introduction to formal semantics, and for a philosophy or logic student wanting a background in formal semantics within linguistics. In the service of being a stand-alone text in modern formal semantic theory, the book will, where relevant, also develop the mechanics of at least one fairly standard non- Direct Compositional theory of the syntax/semantics interface. This is done especially in Part III (Chapters 13–15) where some phenomena are discussed from both direct and non-Direct Compositional points of view. There are several reasons for expositing parallel Direct and non-Direct Compositional accounts of some domains. One is to enable readers to approach the wide range of literature written from either point of view. Second, this allows for a serious comparison of two different approaches. Third, learning more than one set of details for the analysis of any construction allows for a deeper understanding of the basic generalizations and results— generalizations which often transcend the particulars of one theoretical implementation. Finally, a student who has already learned formal seman- tics from a non-Direct Compositional point of view can hopefully also profit from this book by seeing an interesting fragment of English explicitly analyzed from the Direct Compositional point of view. 1.2. A brief note on the history of semantics within modern linguistic theory The subfield of semantics as a core field in modern linguistic theory is relatively recent and is one of the fastest growing subfields.5 Early work within the general enterprise of generative grammar had little to say about semantics. To be sure, by the end of about the 1960s and the early 1970s there was considerable discussion as to how the syntax and the semantics interacted; such discussion was mostly framed in terms of a debate between Generative Semantics (see, e.g., McCawley 1971; Lakoff 1971) and Inter- pretive Semantics (see Chomsky 1970; Jackendoff 1972). We will not discuss the content of that debate here, but much of the work framed within these 5 A much more extensive and authoritative history of the development of formal semantics within modern linguistic theory can be found in Partee (forthcoming). 10 1. INTRODUCTION
two competing points of view did not incorporate a systematic view of the semantics itself. Of course, the linguistics literature during that period contained many seminal observations about semantic notions such as scope, negation, and “binding,” but these were generally not embedded within a full-blown theory of semantics designed to capture semantic notions like entailment and truth conditions (see Chapter 2), although they easily could have been embedded into such a theory. The fact that semantics was not taken a subfield in and of itself during this period comes—at least in part— from Noam Chomsky’s emphasis on syntax during the early development of generative grammar. Chomsky (1957) explicitly rejects the notion that semantics is relevant in the construction of grammars, and this notion persisted for quite some time. It is probably fair to say that modern formal semantics as a subfield within linguistic theory began in the early to mid-1970s with the cross-fertilization of linguistic theory and philosophy of language (including semantics) sparked by Barbara Partee, Richmond Thomason, David Lewis, and others. Par- tee’s work was a particularly influential bridge between linguistics and philosophy as she had originally been a student of Chomsky’s at MIT and always had a strong interest in the connections between language and logic, and hence in topics like quantifiers, negation, etc. As an assistant professor at UCLA, she became acquainted with the seminal work of Richard Mon- tague, a philosopher and logician who (among his many other contributions within philosophy and logic) had a major interest in modeling the semantics (and the syntax/semantics interaction) of natural language (although Montague himself dealt only with English). In fact, the program of Direct Compositionality is advocated in his work (see especially Montague 1970). We will have more to say about his specific contributions as this book proceeds; for now, we note that one of the appeals of his work from the point of view of a linguist was his notion that the semantic composition of natural language reflects and respects its syntax. Partee saw the relevance of Montague’s work to linguistic theory and wrote a series of papers aimed at synthesizing some of the insights from Montague’s work with results within Transformational Grammar (see, for example, Partee 1973). At the same time, the appearance of Lewis (1970), Stalnaker and Thomason (1973), and other work in the philosophy of language also helped launch modern formal semantics and cement its connection to linguistic theory. Such work within philosophy as well as Partee’s early group of students (both at UCLA and later at the University of Massachusetts) continued the tradition, 1.2. THE HISTORY OF SEMANTICS 11
broadening the domain of inquiry and results. From there was born the enterprise known as Montague grammar6 which eventually gave rise to the more general subfield of formal semantics. Montague himself died in 1971,7 and the field of formal semantics evolved in many ways quite different from the original work in Montague grammar. Nonetheless, many of the basic tools of linguistic formal semantics as it is developed to this day stem from some of this early work cited above. Since the late 1970s the field has blossomed, and is now within linguistics generally considered as one of the core areas along with at least phonology and syntax. 1.3. The notion of a “fragment” and its use in this text Inspired by the work of Montague in papers such as Montague (1973), much work in formal semantics within the 1970s and 1980s took it as axiomatic that a goal was to formulate fully explicit grammars (in both syntactic and semantic detail) of the fragment of the language one is concerned with (English in most such work). The term “fragment” got extended to mean not only the portion of the language being modeled, but also the portion of the grammar being proposed as an explicit account of the facts. The strategy of writing fragments (of grammars) has the advantage of giving an explicit theory which makes testable predictions, and of making theory and/or proposal comparison easier. Unfortunately, the goal of formulating fully explicit fragments went out of style during the last two decades or so. This is in part due to the fact that linguistic theories often promised that many of the particular details did not need to be stated as they would fall out from very general principles. It is certainly reasonable to hope that this is ultimately true, but the relevant principles often go unstated or are stated only rather vaguely, making it extremely difficult to really compare proposals and/or evaluate theories and theoretical claims. Having rules and principles be as general as possible is, of course, highly desirable. But this does not mean that they should not be 6 An excellent introduction to the general program of Montague semantics and an explication especially of Montague (1973) can be found in Dowty, Wall, and Peters (1981). 7 Montague was murdered on March 7, 1971. No arrest was ever made in conjunction with the murder. 12 1. INTRODUCTION
formulated explicitly—only that more mileage will be gotten out of explicit formulations. The present text is therefore committed to trying to revive the notion of explicit fragment construction. We cannot promise to give every detail of the domain of English syntax and semantics we are trying to model. Some parts will be left tentative, some stated informally, and some simply omitted. Nonetheless, the goal is to give a reasonable amount of an explicit fragment. We will therefore periodically take stock by summarizing the fragment constructed so far, and a full summary is provided at the end of Part III. *1.4. An intriguing puzzle This introductory chapter concludes with an illustration of a puzzle, a solution to which is proposed in section 15.6. However, the goal here is not to champion any one particular solution, and readers may safely skip this section and return to the data only in section 15.6. But we include this in the introductory remarks for the reader who wants a preview of just what kinds of complex and subtle data a theory of syntax and semantics ultimately hopes to account for. To fully appreciate the particular puzzle here, one should keep the following in mind. The contrasts are quite real; the judgments have been checked with many speakers over the years by myself and many others. Yet—like other subtle facts in syntax, phon- ology, and semantics—these are not generalizations which we have ever been consciously taught nor even generalizations that most of us are even aware of until we see them in a linguistics course (or book). What, then, is there about our unconscious knowledge of the grammatical system that predicts these judgments? This is the sort of puzzle that theories of semantics and its interaction with syntax ultimately seek to solve. So, consider what we will call the A-B-C party scenario. I go to a small party consisting of only myself and three married couples: Alice and Abe, Betty and Bert, and Cathy and Carl. I learn that Alice and Abe met each other only a few years ago, and similarly for Cathy and Carl. But interest- ingly, I also find out that Betty and Bert have been sweethearts since childhood. I like Betty a lot, and spend a good part of the evening talking to her. The next day, you ask me how I enjoyed the party and if there was anyone that I especially enjoyed meeting. I certainly can answer with (3): *1.4. AN INTRIGUING PUZZLE 13
(3) Oh yes, I especially enjoyed talking to Betty. But now suppose that I can’t remember Betty’s name, although I do remem- ber that her husband’s name is Bert. I can answer with either (4a) or (4b): (4) a. Oh yes, I especially enjoyed talking to—oh, I can’t remember her name— you know, the woman who is married to Bert. b. Oh yes, I especially enjoyed talking to—oh, I can’t remember her name— you know, the wife of Bert. ((4b) would sound more natural if we substituted Bert’s wife for the wife of Bert; this will not impact on the ultimate point and the exposition is simplified using (4b).) Now, let us tweak the scenario slightly and assume that I am one of those people who just doesn’t remember names very well. As a result, I remember neither Betty’s name nor Bert’s name, although I do remember the interest- ing fact that they are the only couple at the party who have been sweethearts since childhood. As an answer to your question, (5) would be quite natural: (5) Oh yes, I especially enjoyed talking to—oh, I can’t remember her name—you know, the woman who is married to her childhood sweetheart. But what is striking is that I can’t answer with (6): (6) *Oh yes, I especially enjoyed talking to—oh I can’t remember her name—you know, the wife of her childhood sweetheart. (We are taking liberties with the * notation here. This is generally used in works in syntax to indicate something that is ill-formed. (6) in fact is fine, just not on the intended reading; and we will continue to notate a sentence with an asterisk in front of it when we mean “bad on a particular reading” provided that it is clear what the intended reading is.) It should be noted that some speakers find the contrast rather subtle but there is general agreement that (6) is stranger than (5). All of these examples contain various extra material (the parentheticals, etc.) which are there to make them sound natural and conversational. But as we proceed it will be convenient to strip away the parts that are irrelevant to figuring out the semantics, so we can recast (6) as (7)—also impossible as an answer to the question and in this context: (7) *I especially enjoyed talking to the wife of her childhood sweetheart. 14 1. INTRODUCTION
Of course, (7) is a perfectly good sentence, but it cannot be used in our party scenario as a way to identify Betty. Since some readers do find the contrast subtle, two points are worth noting. First, one should resist the temptation to recast (7) in one’s mind as I especially enjoyed talking to the one who’s the wife of her childhood sweetheart or I especially enjoyed talking to the woman who’s the wife of her childhood sweetheart. That would be cheating; the point is not to find a closely related way to say the same thing but to notice that the actual way in (7) contrasts with I especially enjoyed talking to the woman who is married to her childhood sweetheart (and contrasts with the above variants too). As to why these variants are good, we return to that shortly. Moreover, while the contrasts above may be subtle for some speakers, there is a related mystery where the facts dramatically pop out. Thus take (8) in the same scenario, where the only people at issue are Alice, Betty, and Cathy: (8) Betty is the only woman who is married to her childhood sweetheart. This can be making two different claims. The obvious one in this scenario is that Cathy is not married to Cathy’s childhood sweetheart, and Alice is not married to Alice’s childhood sweetheart. The other is the “non-polygam- ous” reading: it asserts that Bert (or whoever Betty’s husband might be) has only one wife. Since we (generally) assume that people have just one wife, this reading (given standard assumptions) is not the first one that someone would think of since it is less likely to be conveying any interesting information. But despite the fact that the non-polygamous reading is the less obvious one for (8), it is the only reading (or at least the one that pops out first) for (9): (9) Betty is the only wife of her childhood sweetheart. Why should that be? We’ll put (8) and (9) aside for the moment, and return to the simpler case of (7). So the mystery is why (7) is bad as a way to identify Betty. This is especially puzzling in that both (4b) and (5) are perfectly good—or, to give their stripped-down versions, (10) and (11) are both fine. Each one differs minim- ally from our bad case, yet neither of these two has any problem. (10) I especially enjoyed talking to the woman who is married to her childhood sweetheart. (11) I especially enjoyed talking to the wife of Bert. *1.4. AN INTRIGUING PUZZLE 15
So surely there is nothing incoherent or wrong with the meaning that (7) is trying convey, for (10) is just a slightly different form and conveys exactly this meaning. Hence the puzzle has something to do with the mapping between syntax and semantics: why one is a good way to package the relevant information while the other is not. We can informally recast the puzzle in the following way. Compare the two expressions the woman who is married to Bert and the wife of Bert. (Following a long tradition within linguistics, we will refer to these as NPs, which comes from “noun phrases.” They are also in much modern literature referred to as DPs, for “determiner phrases,” but we stick to the more traditional terminology in this text.)8 Both of these can correspond to meanings that we can (roughly and informally) represent as (12): (12) the x: x is a woman and x is married to Bert But while the object NP in (10) can be represented as in (11), the object NP in (7) cannot: (13) the x: x is a woman and x is married to x’s childhood sweetheart The basic phenomenon here was discussed in, among others, Jacobson (1977) (where it was called Langendoen’s constraint), Chomsky (1981 under the rubric of i-within-i condition), and many since. As there seems to be nothing wrong with the meaning, we can assume that the phenomenon in question has something to do with the way the syntax and semantics interact. Notice that we have given a kind of formula (and one that uses a “variable” x) to represent the meanings in question, but for now we should think of these simply as placeholders to bring out the intuition. After all, recasting her in the above examples as x doesn’t really immediately give us the tools for computing the meanings of the expression: we have traded a 8 In theories which use the term DP, the NP is used instead to refer to material after the Determiner; e.g., mother of Romeo in an expression like the mother of Romeo. Here we will be calling this simply N (i.e., a “noun”) and allowing terms like N to refer both to simple material consisting of just one word and to complex material. This is discussed further in Chapter 6. We are aware that this will initially cause some confusion to a reader who is used to using “NP” to mean a noun and its complement, but it is well worth becoming fluent in both sets of terminologies. The terminology here is the traditional one found in large amounts of literature not only in syntax and semantics but in neighboring fields like psycholinguistics, philosophy, cognitive science, etc. 16 1. INTRODUCTION
pronoun her for a variable x. But this accomplishes little until we have a way to think about what a variable like x means. (Indeed this is explored in detail in Chapters 9 and 15, including developing an alternative view that does not make use of variables in the semantics.) We thus caution that formulas like (12) and (13) are best seen simply as informal and helpful ways to bring out the intended meanings. Similarly, one often sees indices used in the literature as a way to bring this out; one will find discussions using the notation in (14) and (15) to make the point, where the indexation in (14) indicates a good possible reading for the NP while (15) cannot be understood in the intended way: (14) the womani whoi is married to heri childhood sweetheart (15) *the wifei of heri childhood sweetheart Much work in grammatical theory actually assumes that NPs and pronouns come with (obviously silent) indices in the syntax; here we will be using indices from time to time simply as a way to notate intended readings without any commitment to their being actual pieces of grammatical machinery. Before leaving this (for now), there’s one other interesting point to notice. However we ultimately state the principle, the claim is that an NP like the wife of her childhood sweetheart cannot correspond to the meaning shown earlier in (13): (13) the x: x is a woman and x is married to x’s childhood sweetheart But one might think that this is not really correct, since it is in fact just fine to use (16) as a way to identify Betty: (16) I especially enjoyed meeting the woman who is the wife of her childhood sweetheart. This point was made earlier; many speakers on reading (7) tend to recast it in their minds as (16). Similarly, (17) is impeccable on the understanding where her is Betty: (17) Betty is the wife of her childhood sweetheart. But a close reflection reveals that this does not threaten the generalization. Again, using indices or variables simply as a convenient way to elucidate the point, it is easy to see in (17) that her need not be “coindexed with” (or “correspond to the same variable as”) wife but rather it just refers directly to Betty. That is, we can represent it as in (18a) using indices, or as in (18b) using the more spelled-out formula (though still quite informal). *1.4. AN INTRIGUING PUZZLE 17
(18) a. Bettyj is [NP the wifei of herj childhood sweetheart]. b. Betty, y [y = the x: x is a woman and x is married to y’s childhood sweetheart] Since we are asserting identity between Betty and the person married to Betty’s childhood sweetheart, it of course follows that Betty is married to Betty’s childhood sweetheart and so the full sentence (16) will end up with the relevant meaning.9 But the claim that the object NP itself (the wife of her childhood sweetheart) does have the meaning represented in (12) is not threatened. The same point holds for (17), whose meaning can be represented as (19a) or (19b). (19) a. the womanj whoj is the wifei of herj childhood sweetheart b. the y: y is a woman and y = the x: x is a woman and x is married to y’s childhood sweetheart Is there a way to confirm that this is the right sort of explanation for these apparent counterexamples? Indeed there is, and it centers on the contrast between (8) and (9) which was discussed earlier. We leave it to the interested reader in the exercise to play with this and get a sense of why (8) is ambiguous and (9) is not. Having completed that, one should be able to see how it is that this gives support for the explanation offered above as to why (17) does not threaten the claim that the wife of her childhood sweetheart cannot correspond to the meaning shown informally in (12). *1.2. Work out—using the informal representations either with indices or the representations with variables—why it is that (7) is ambiguous and (8) is not. Of course you will need to think a bit about how to treat only, but nothing very complex is required. You can be perfectly informal in your treatment of only, but you should be able to get a feel for why these two differ. 9 This general observation—although for a slightly different case—was made in Postal (1970) who distinguished between “presuppposed” coreference and “asserted” coreference. Here the fact that Betty and the wife of her childhood sweetheart end up “referring” to the same individual is exactly what the sentence is asserting. 18 1. INTRODUCTION
As noted at the outset of this section, the goal here is just to provide a mystery to whet the reader’s appetite; the tools needed to provide a hypoth- esis as to the explanation of the mystery will be developed later. 1.5. Appendix: Sets and Functions 1.5.1. Sets, members, and subsets Since the notions of sets and of functions are crucial throughout this book, some formal definitions and discussion are provided here for readers not entirely familiar with these notions. We begin with the notion of a set. A set is simply any collection of objects (it can have a finite number of objects, an infinite number, or none at all). For example, we can talk about the set of positive integers less than 10; sets can be notated by listing the members and enclosing the list in curly brackets: {1,2,3,4,5,6,7,8,9}. The order in which they are listed makes no difference; a set is just a collection of things without any order. So if we were to write {2,5,3,4,9,7,8,1,6}, this names the same set. Each integer in this set is called a member or an element of the set. If we were to name this set A, then the notation 4 ∈ A means that 4 is a member (or element) of A. Something either is or is not in a set; it makes no sense to say it occurs twice (or more) in the set. Note also that a set can have a single member; this is called a singleton set. Thus {4} is the set with only one member; this set is distinct from 4 itself. (4 is a member of {4}.) A set can have an infinite number of members; the set of positive integers for example is infinite. Obviously this can’t be named by listing the mem- bers. One can in this case specify the set by a recursive procedure. Call the set I, then one can specify I by two statements: (a) (what is known as the base step): 1 ∈ I, and (b) (the recursion step) if n ∈ I then n+1 ∈ I. (It is understood when one lists things this way that nothing else is in I.) One will also often see a notation which describes rather than lists the members. For example, we can write the following set, call it B: {x|x is a New England state}. This names a finite set, and so we could also give B in list form as follows: {Maine, New Hampshire, Vermont, Massachusetts, Rhode Island, Con- necticut}. These two are just different notations for naming the same set. This can also be used, of course, for infinite sets. Take, for example, the set 1.5. APPENDIX: SETS AND FUNCTIONS 19
{x|x is an integer and x > 9}. This names the set of integers greater than 9. And, a set can have no members. There is only one such set; its name is the null set or the empty set, and is generally written as . Of course, there are other ways one can describe the null set. For example, the set of integers each of which is greater than 9 and less than 10 is the empty set. The cardinality of some set refers to the number of elements in that set; the notation |B| means the cardinality of B. Hence, given our set B above, |B| is six. Take some set A. Then a subset of A is any set all of whose members are also in A. Suppose, for example, we begin with a set C which is {1,2,3}. Then {1,2} is a subset of C, as is {1,3} and so forth. The notation for the subset relation is . The full definition of subset is as follows: B A if and only if every member of B is a member of A. From this it follows that every set is a subset of itself (so for the set C above, one of its subsets is the set {1,2,3}). It is, however, sometimes convenient to refer to those subsets distinct from the original set; in that case we can talk about a proper subset of some set. The symbol for this is , so B ` if and only if B A and B 6¼ A. Since the definition of subset says that B is a subset of A if and only if everything that is in B is also in A, it follows that if nothing is in B then B is a subset of A. thus the null set is a subset of every other set. Sets themselves can have sets as members, and so one can talk about the set of all subsets of a set A. This is called the power set of A, written as P (A). For example, given the set C above, P (A) = {, {1}, {2}, {3}, {1,2}, {1,3}, {2,3}, {1,2,3}}. We will also have occasion to talk about the reverse of the subset relation—i.e., the superset relation. A is a superset of B if and only if B is a subset of A. The notation for this is A B. Once again this is defined in such a way that every set is a superset of itself; a superset of B which is not identical to B is called a proper superset, and the notation for this is . *1.3. If a set A has n members, then the number of subsets of A is 2n . Try to see why this is true. Hint: for every member x of some set A, then for each subset B of A, x is either in B or is not in B. 1.4. How many members does the following set have: {}? *1.5. What is P ()? 20 1. INTRODUCTION
1.5.2. Union, intersection, and complement Take any two sets A and B. Then there is a set C which consists of everything that is in A and everything that is in B. This is called the union of A and B, and is written A [ B. For example, if A is {1,2,3} and B is {2,4,6} then A [ B is {1,2,3,4,6}. Moreover, for any two sets A and B the intersection of A and B is the set of all things that are in both A and B. This is written A B. So, for example, in the case directly above, the intersection of A and B is {2}. Or, if we were to intersect the set of integers which can be evenly divided by 2 (the set of even integers) with the set of integers which can be evenly divided by 3, we end up with the set of integers that can be evenly divided by 6. One final useful notion here is the complement of a set. The complement of some set A is the set of all things which are not in A (this is sometimes notated as A0 ). Usually one talks about this notion with respect to some larger domain. Strictly speaking, the complement of {1,2,3} would include not only all integers greater than 3 but also all sorts of other numbers (like 1/3), the sun, my dog Kiana, and the kitchen sink. Rarely are we interested in that sort of set; so in practice when one talks about “the complement of some set A” this is generally with respect to some larger set B of which A is a subset. Then the complement of A refers to all things in B that are not in A (this is notated as B-A). For example, when restricting the discussion to the set of positive integers, the complement of {1,2,3} is the set of all integers greater than 3. 1.5.3. Ordered pairs, relations, equivalence relations, and partitions Sets are unordered collections of objects. But it is quite useful (as will become very apparent as this book proceeds) to be able to talk about pairs of objects that are ordered in some way. An ordered pair is just that: 1.6. a. For any two sets A and B such that A B, what set is A [ B? b. For any two sets A and B such that A B, what set is A B? 1.5. APPENDIX: SETS AND FUNCTIONS 21
it is two objects with some ordering between them. If the two objects are a and b, then (a,b) is an ordered pair; (b,a) is a different ordered pair. An ordered pair need not contain distinct items: (a,a) is an ordered pair. In applying this to actual natural relations that exist in the world we are generally interested in sets of ordered pairs. (One can generalize this notion to ordered triples and so forth; an ordered n-tuple means an ordered list of n items.) This notion is easiest to grasp with some concrete examples. Take again the set {1,2,3}, and take the relation “is greater than.” Then this can be seen as a set of ordered pairs; if we are restricting this to items from our little 1-2-3 set, this would be the set {(2,1), (3,1), (3,2)}. Now suppose we instead take the following set of ordered pairs: {(2,1), (3,1), (3,2), (1,1), (2,2), (3,3)}. Then (restricting this again to our 1-2-3 set) we have now actually listed the relation “is greater than or equal to.” Or, take the set {(1,1), (2,2), (3,3)}. That is the relation “is equal to” (defined for the set of integers {1,2,3}). In other words, what we are calling a relation is just some set of ordered pairs. In the example above, both the first and second member of each ordered pair was drawn from the same set (the set {1,2,3}). But this is not necessary; we can have a set of ordered pairs each of whose first member is drawn from some set A and the second member from some set B where A and B are different (they can, but need not, have some of the same members). For example, the relation “is the capital of” is a relation between cities and states; it can be expressed as a set of ordered pairs of the general form {(Providence, Rhode Island), (Boston, Massachusetts), (Springfield, Illinois), (Pierre, South Dakota), . . . } (the . . . here is a shorthand for the remaining 46 pairs). Take two sets A and B (they could be the same set or different). Then A x B refers to the set of all ordered pairs whose first member is in A and whose second member is in B. (This is also called the Cartesian product of A and B.) As in the case above, it is helpful to give the intuition of this by coming up with some concrete example. Suppose we take as our set A some group of professors—say, Professor Magoo, Professor Carberry, and Professor Glazie. Call that set P (for shorthand, let’s call its members m, c, and g, so P is the set {m,c,g}). Now suppose we have a set S which consists of three students who we will just indicate as x, y, and z (so S = {x,y,z}). Then P x S = {(m,x}, (m,y), (m,z), (c,x), (c,y), (c,z), (g,x), (g,y), (g,z)}. Suppose that Magoo wrote a letter of recommendation for all three students, Carberry wrote one for only y, and Glazie wrote one for y and z. Then the relation “wrote a 22 1. INTRODUCTION
recommendation for” is a subset of P x S, and is the set of ordered pairs {(m,x}, (m,y), (m,z), (c,y), (g,y), (g,z)}. More generally, we define a relation (between members of A and B) as any subset of A x B. There are some special and interesting properties that can be defined when the two sets are the same. That is, we are now looking at subsets of A x A. Consider a relation R (some subset of A x A) which is such that for all x in A, (x,x) is in R. Such a relation is called a reflexive relation. (These need not be the only kinds of pairs to be in R for R to be reflexive; other pairs can be in there too.) For example, if talking about the set of integers again, the relation “is greater than or equal to” is reflexive; for all numbers n, (n,n) is in the set of ordered pairs described by that relation. A relation R is called irreflexive if for all x in A, (x,x) is not in R. Further, consider any two members x and y (both members of A). Then if it’s the case that for all x and y if (x,y) is in R then (y,x) is also in R, the relation is called symmetric. Imagine, for example, a lovely world with no unrequited love. Then is in love with is symmetric in that world. If our set were {m, c, g, and p}, then if the pair (m,c) were in our relation R (i.e., “is in love with”) the fact that R is symmetric means that (c,m) is also in R. (Notice that our definition neither requires (c,c) to be in R nor excludes that; either is possible.) Or, to look at a relation which is symmetric by definition: consider the relation is a sibling of. (While is a sibling of is symmetric, is a sister of is not. Why not?) One final useful definition is a transitive relation. A transitive relation R is one for which for every x, y, and z, if (x,y) is in R and (y,z) is in R, then (x,z) is in R. (The relation “is greater than” is transitive, as is the relation “is greater than or equal to”). Any relation R which is reflexive, transitive, and symmetric is called an equivalence relation. As an example of such a relation, consider the set of students (call it S) at an elementary school that services grades 1 through 6. Then “is in the same grade as” is an equivalence relation in S x S. (While it is unusual to use the phrase “in the same grade as” when referring to the same person it seems false to say Johnny is not in the same grade as himself so we can see that this relation is reflexive.) It is also obvious that it is symmetric and transitive. Note that this—and any other equivalence relation—divides up the original set (here, S) into a group of non-overlapping subsets. The set of these subsets is called a partition. Thus, a partition of any set S is a set of subsets of S such that for each distinct subset A and B, A B = , and the union of all the subsets is S. To show that any equivalence relation induces such a partition, take any x in S and define Sx as {y|(y,x) is in R}. Since R is 1.5. APPENDIX: SETS AND FUNCTIONS 23
reflexive, we know that x is in Sx (and hence we know that Sx is guaranteed not to be empty). Moreover, the fact that R is reflexive means that each member of S is guaranteed to be in at least one such subset, so we know that the union of all of these is S. We can further show that for any two such subsets Sa and Sb, they either have no members in common (i.e., they have a null intersection) or they are the same. Thus, take any c which is in both Sa and Sb. By definition, this means that (c,a) is in R and (c,b) is in R. By the fact that R is transitive and symmetric, it follows that (a,b) and (b,a) are in R (the reader can work through the necessary steps). But then, for all x such that (x,a) is in R, (x,b) is also in R. To show this note again that R is transitive. If (x,a) is in R and (a,b) is in R then (x,b) is also in R. Hence given the initial premise that there is a non-empty intersection between Sa and Sb, it follows that everything in Sa is in Sb. That everything in Sb is also in Sa follows in the same way, and so the two are the same set. Each subset in a partition is called a cell in that partition. In the example above, the cells correspond to the different grades. (There don’t have to be six cells—it could be that one of the grades has no student in it. But there can be no more than six; recall that by definition a cell can’t be empty.) Just as any equivalence relation induces a partition, given any partition one can give an equivalence relation that corresponds to any partition; this is the relation that holds between any two a and b in S such that a and b are in the same cell in the partition. 1.5.4. Functions A function takes every member of some set A and assigns it a value from a set B (B could be the same set as A, but need not be). This can also be formalized using the notion of a set of ordered pairs. Thus, consider two sets A and B (which again could be the same but need not be). Then, a (total) function from A to B is any set of ordered pairs (i.e., any subset of A x B) such that for each a in A, there is one and only one ordered pair with a as first member. Thus if we think of the function f as assigning to each a in A some values in B, note that the criterion above ensures that each member of A is indeed assigned a value, and is assigned a unique value. A is referred to as the domain of the function, and B is referred to as the co-domain. For any function f and any a in the domain of f, we write f(a) to indicate the value that f assigns to a. (To use other common terminology, f(a) means 24 1. INTRODUCTION
the result that one gets by applying the function f to a.) There is no restriction that each member of B must appear as second member of some ordered pair; the term range of the function f is the set of all b in B such that there is some a such that f(a) = b. Note that these definitions are such that the range of a function is a subset of the co-domain. In practice (at least in works within linguistics) the terms “range” and “co-domain” are often not distinguished. As noted above, there is no restriction that each member of B appear as second member of an ordered pair. Nor is there a restriction that it appear only once. If each member of B is used as a value only once (that is, for each b in B, there is a unique a such that f(a) = b) then B obviously can be no smaller than a. It can have more members, or it can be the same size. If the latter is the case, then it also follows that for every b in B, there is some a such that f(a) = b. When both conditions above hold (i.e., for each b in B, there is one and only one a such that f(a) = b, we say that there is a one-to- one correspondence between A and B. Note that for any function f which is a one-to-one correspondence, there is a corresponding function f-1 which is just the reverse: it is a function mapping each member of B to a member of A such that for all a in A and b in B, if f(a) = b then f-1 (b) = a.10 We will have some occasion to talk about the notion of a partial function. A partial function is one where not every member of A is actually assigned a value by f; f is undefined for some subset of A. (Of course any partial function f is also a total function with a smaller domain.) We can illustrate this by returning to our earlier example of ordered pairs of US cities and states, where the first member of each ordered pair is the capital of the second. This is a partial function from the set of US cities to states (not every US city is a capital). We can reverse it, and have each state as the first member of the ordered pair and the second as its capital (this function could be expressed in prose as has as its capital). This is now a total function from 10 Incidentally, the notion of the availability of a one-to-one correspondence can be used to define what it means for two sets to have the same cardinality. Obviously for two finite sets it is clear what it means to have the same cardinality, since we can count the members. But consider the case of infinite sets. Take the following two sets: A = the set of positive integers {1,2,3, . . . } and B = the set of positive even integers {2,4,6, . . . }. Both are infinite. Surprisingly (when one first hears this) they are also of the same cardinality, because one can establish a one-to-one correspond- ence between them (each member of A is paired with a member of B by multiplying by 2: we will never run out of members in B). 1.5. APPENDIX: SETS AND FUNCTIONS 25
the set of states (every state does have a capital) to the set of US cities. But it is not a one-to-one correspondence for the same reason that our original relation is not a total function; there are many cities without the honor of being a capital. Occasionally in this text it will be useful to list out some actual functions—that is, to name every member in the domain and name what the function at issue maps that member to. There are a variety of ways one could do this. To illustrate, take a domain of four children {Zacky, Yonnie, Shelley, and Baba} (call that set C) and four men {Abe, Bert, Carl, David} (call that set M). Suppose there is a function f from C to M which maps each child to their father. Assume that Abe is the father of Zacky and Yonnie, Bert is the father of Shelley, and David is the father of Baba. Then one can write this information out in various ways. One would be to simply give the set of ordered pairs: {(Zacky, Abe), (Yonnie, Abe), (Shelley, Bert), (Baba, David)}. Usually this notation, however, is not terribly easy to read. We could also write this out in either of the ways shown in (20): (20) a. f(Zacky) = Abe b. Zacky ! Abe f(Yonnie) = Abe Yonnie ! Abe f(Shelley) = Bert Shelley ! Bert f(Baba) = David Baba ! David Or, sometimes it is more convenient to list out the domain on the left and the co-domain on the right and connect them with arrows as in (21): (21) Zacky Abe Yonnie Bert Shelley Carl Baba David Which notation is chosen makes no difference; the choice should be dictated by clarity. 26 1. INTRODUCTION
2 Semantic foundations 2.1. Model-theoretic semantics 27 2.2. Truth conditions 28 2.3. Possible worlds 31 2.4. Times 41 2.1. Model-theoretic semantics The primary focus of this book is the syntax/semantics interface—that is, how the syntax and semantics work so that sentences (and other well- formed linguistic expressions) are paired with a meaning. But of course this task is impossible without some idea of what meaning is. Can we talk about meaning without relegating it to the realm of the mysterious, or leaving it solely to folks who work on cognition to deal with? The answer (of course) is yes—there is a rich tradition within linguistics and the philosophy of language for modeling linguistic meaning. In some early work within linguistic theory—especially in the 1960s— meaning was taken to be just a symbolic representation (call it a Logical Form, or LF). While it remains an open question as to whether such representations play an actual role in the way the grammar pairs expressions with meanings, this book (along with much other modern work in formal semantics) assumes that meaning is not just some string of symbols, but rather some actual object out there in the world. Call this a model-theoretic object. (More precisely, we are taking meaning to be an object which forms part of a model which is an abstract representation of the world: hence the term model theory.) Of course, we need some way to name these objects, and so throughout we will use strings of symbols as ways to name them. But the
point is that the grammar maps each linguistic expression into something beyond just a symbolic representation. Otherwise—as so aptly pointed out by David Lewis (1970)—we are simply mapping one language (say, English) into another (what Lewis termed “Markerese”). Yet language is used to convey facts about the world; we draw inferences about the world from what we hear and we gain information about what is true and what is not. So semantics must be a system mapping a linguistic expression to something in the world. But what exactly is meant by model-theoretic objects? These can in fact be quite abstract. Still, they are the “stuff” that is out there in the universe— something constructed out of actual bits of the universe (or, at least, the ontology of the universe as given by language). This would include things like individuals, times, possibilities, and perhaps others; just what are the basic objects that we need is an open question and is part of what semantic theory addresses. The strategy here will be to use a fairly sparse set of primitive objects, and construct more complex objects out of these. Let us, then, begin by setting up two basic building blocks which are foundational in much of the work in linguistic formal semantics. 2.2. Truth conditions A common adage in semantics is: “To know the meaning of a (declarative) sentence is to know what it would take to make it true.” We can use this adage as a first step in constructing the building blocks for meanings: a fundamental fact about declarative sentences is that they are either true or false1 (and since we use language to communicate information about the world, a listener will in general assume that a sentence they have just heard 1 Henceforth we use the term “sentence” to mean a declarative sentence. There is actually no reason to consider questions to be of the same category as declarative sentences even though they also are traditionally referred to as “sentences.” Ques- tions have a different external distribution from declarative sentences (for example, wonder can occur only with a question as its complement, not an ordinary sentence, while the reverse is true for believe); they have a different kind of meaning, and they have a different internal structure. Whether imperatives and declaratives should be considered the same category is a bit less clear (they are more similar), but we will not deal with those here either. 28 2. SEMANTIC FOUNDATIONS
is true, and uses that fact to enrich their knowledge of the world). Thus (1) is true and (2) is false: (1) Barack Obama moved into the White House on Jan. 20, 2009. (2) John McCain moved into the White House on Jan. 20, 2009. Hence, one basic notion used for the construction of meanings is a truth value—for now assume that there are just two such values: true and false. (More on this directly.) The claim that truth values are a fundamental part of meaning is also motivated by noting that—as shown by the examples above—speakers have intuitions about truth, given certain facts about the world, just like they do about acceptability. And these judgments can be used to test the adequacy of particular theories of meaning. Following standard practice, we use 1 for true and 0 for false. Thus the set of truth values {1,0} and we will also refer to this set as t. Let us use [[Æ]] to mean the semantic value (i.e., the meaning) of a linguistic expression Æ. Then (temporarily) we can say that [[Barack Obama moved into the White House on Jan. 20, 2009]] = 1. Some worries should immediately spring to mind. The most obvious is that something seems amiss in calling the meaning of (1) “true” even if we are willing to accept the fact that it is true. We will enrich the toolbox directly to take care of that. But there are other objections: does it really make sense to say that all declarative sentences are true or false? Clearly not—for some sentences the truth value depends on who is speaking (and on when the sentence is spoken). Take (3): (3) I am President of the United States. This is true if spoken by Barack Obama in 2011, but not if spoken by John McCain and not true if spoken by Barack Obama in 2006. So this has no truth value in and of itself. Nonetheless once certain parameters are fixed (time of utterance and speaker) it is either true or false. So we might want to think of the meaning of (3) as a function into {1,0}—it does yield a truth value but only once we fix certain parameters. But it seems inescapable that a declarative sentence is telling us something about the world, and so truth values are certainly one fundamental piece. In fact, there are many parameters that need to be set in order to derive a truth value. Certain words like I, you, here, now, etc. quite obvi- ously have the property that their value depends on when, where, and by whom these are spoken (these are called indexicals). There are also more 2.2. TRUTH CONDITIONS 29
Another Random Document on Scribd Without Any Related Topics
of such an animal shows an abundant vegetation, and a climate so mild, that the rivers were not covered with heavy ice in winter; for the supposition that this old hippopotamus was a migratory animal seems very unlikely. Another animal of this time, was the magnificent deer, known as the Irish elk; and which perhaps had its principal abode on the great plain which is now the Irish Sea. The terrible machairodus, or cymetar-toothed tiger, was continued from the Pliocene; and in addition to species of bear still living, there was a species of gigantic size, probably now extinct, the cave bear. Evidences are accumulating, to show that all or nearly all these survived until the human period. If we turn now to those animals which are only locally extinct, we meet with some strange, and at first sight puzzling anomalies. Some of these are creatures now limited to climates much colder than that of Britain. Others now belong to warmer climates. Conspicuous among the former are the musk-sheep, the elk, the reindeer, the glutton, and the lemming. Among the latter, we see the panther, the lion, and the Cape hyena. That animals now so widely separated as the musk-sheep of Arctic America and the hyena of South Africa, could ever have inhabited the same forests, seems a dream of the wildest fancy. Yet it is not difficult to find a probable solution of the mystery. In North America, at the present day, the puma, or American lion, comes up to the same latitudes with the caribou, or reindeer, and moose; and in Asia, the tiger extends its migrations into the abodes of boreal animals in the plains of Siberia. Even in Europe, within the historic period, the reindeer inhabited the forests of Germany; and the lion extended its range nearly as far northward. The explanation lies in the co-existence of a densely wooded country with a temperate climate; the forests affording to southern animals shelter from the cold or winter; and equally to the northern animals protection from the heat of summer. Hence our wonder at this association of animals of diverse habitudes as to climate, is merely a prejudice arising from the present exceptional condition of Europe. Still it is possible that changes unfavourable to some of these animals, were in progress before the arrival of man, with his clearings and forest fires and other disturbing agencies. Even in America, the megalonyx, or gigantic sloth, the mammoth, the mastodon, the fossil horse, and many other creatures,
disappeared before the Modern period; and on both continents the great Post-glacial subsidence or deluge may have swept away some of the species. Such a supposition seems necessary to account for the phenomena of the gravel and cave deposits of England, and Cope has recently suggested it in explanation of similar storehouses of fossil animals in America.[AS] [AS] Proceedings of the American Philosophical Society, April 1871. Among the many pictures which this fertile subject calls up, perhaps none is more curious than that presented by the Post-glacial cavern deposits. We may close our survey of this period with the exploration of one of these strange repositories; and may select Kent’s Hole at Torquay, so carefully excavated and illumined with the magnesium light of scientific inquiry by Mr. Pengelly and a committee of the British Association. The somewhat extensive and ramifying cavern of Kent’s Hole is an irregular excavation, evidently due partly to fissures in limestone rock, and partly to the erosive action of water enlarging such fissures into chambers and galleries. At what time it was originally cut we do not know, but it must have existed as a cavern at the close of the Pliocene or beginning of the Post-pliocene period, since which time it has been receiving a series of deposits which have quite filled up some of its smaller branches. First and lowest, according to Mr. Pengelly, is a “breccia” or mass of broken and rounded stones, with hardened red clay filling the interstices. Most of the stones are of the rock which forms the roof and walls of the cave, but many, especially the rounded ones, are from more distant parts of the surrounding country. In this mass, the depth of which is unknown, are numerous bones, all of one kind of animal, the cave bear, a creature which seems to have lived in Western Europe from the close of the Pliocene down to the modern period. It must have been one of the earliest and most permanent tenants of Kent’s Hole at a time when its lower chambers were still filled with water. Next above the breccia is a floor of “stalagmite” or stony carbonate of
lime, deposited from the drippings of the roof, and in some places three feet thick. This also contains bones of the cave bear, deposited when there was less access of water to the cavern. Mr. Pengelly infers the existence of man at this time from a single flint flake and a single flint chip found in these beds; but mere flakes and chips of flint are too often natural to warrant such a conclusion. After the old stalagmite floor above mentioned was formed, the cave again received deposits of muddy water and stones; but now a change occurs in the remains embedded. This stony clay, or “cave earth” has yielded an immense quantity of teeth and bones, including those of the elephant, rhinoceros, horse, hyena, cave bear, reindeer, and Irish elk. With these were found weapons of chipped flint, and harpoons, needles, and bodkins of bone, precisely similar to those of the North American Indians and other rude races. The “cave earth” is four feet or more in thickness, It is not stratified, and contains many fallen fragments of rock, rounded stones, and broken pieces of stalagmite. It also has patches of the excrement of hyenas, which the explorers suppose to indicate the temporary residence of these animals; and in one spot, near the top, is a limited layer of burnt wood, with remains which indicate the cooking and eating of repasts of animal food by man. It is clear that when this bed was formed the cavern was liable to be inundated with muddy water, carrying stones and other heavy objects, and breaking up in places the old stalagmite floor. One of the most puzzling features, especially to those who take an exclusively uniformitarian view, is, that the entrance of water-borne mud and stones implies a level of the bottom of the water in the neighbouring valleys of about 100 feet above its present height. The cave earth is covered by a second crust of stalagmite, less dense and thick than that below, and containing only a few bones, which are of the same general character with those below, but include a fragment of a human jaw with teeth. Evidently, when this stalagmite was formed, the influx of water-borne materials had ceased, or nearly so; but whether the animals previously occupying the country still continued in it, or only accidental bones, etc., were introduced into the cave or lifted from the bed below, does not appear.
The next bed marks a new change. It is a layer of black mould from three to ten inches thick. Its microscopic structure does not seem to have been examined; but it is probably a forest soil, introduced by growth, by water, by wind, and by ingress of animals, at a time when the cave was nearly in its present state, and the surrounding country densely wooded. This bed contains bones of animals, all of them modern, and works of art ranging from the old British times before the Roman invasion up to the porter-bottles and dropped halfpence of modern visitors. Lastly, in and upon the black mould are many fallen blocks from the roof of the cave. There can be no doubt that this cave and the neighbouring one of Brixham have done very much to impress the minds of British geologists with ideas of the great antiquity of man, and they have, more than any other Post-glacial monuments, shown the persistence of some animals now extinct up to the human age. Of precise data for determining time, they have, however, given nothing. The only measures which seed to have been applied, namely, the rate of growth of stalagmite and the rate of erosion of the neighbouring valleys, are, from the very sequence of the deposits, obviously worthless; and the only apparently available constant measure, namely, the fall of blocks from the roof, seems not yet to have been applied. We are therefore quite uncertain as to the number of centuries involved in the filling of this cave, and must remain so until a surer system of calculation is adopted. We may, however, attempt to sketch the series of events which it indicates. The animals found in Kent’s Hole are all “Post-glacial.” They therefore inhabited the country after it rose from the great Glacial submergence. Perhaps the first colonists of the coasts of Devonshire in this period were the cave bears, migrating on floating ice, and subsisting, like the Arctic bear, and the black bears of Anticosti, on fish, and on the garbage cast up by the sea. They found Kent’s Hole a sea- side cavern, with perhaps some of its galleries still full of water, and filling with, breccia, with which the bones of dead bears became mixed. As the land rose, these creatures for the most part betook themselves to lower levels, and in process of time the cavern stood
upon a hill-side, perhaps several hundreds of feet above the sea; and the mountain torrents, their beds not yet emptied of glacial detritus, washed into it stones and mud and carcases of animals of many species which had now swarmed across the plains elevated out of the sea, and multiplied in the land. This was the time of the cave earth; and before its deposit was completed, though how long before, a confused and often-disturbed bed of this kind cannot tell, man himself seems to have been added to the inhabitants of the British land. In pursuit of game he sometimes ascended the valleys beyond the cavern, or even penetrated into its outer chambers; or perhaps there were even in those days rude and savage hill-men, inhabiting the forests and warring with the more cultivated denizens of plains below, which are now deep under the waters. Their weapons, lost in hunting, or buried in the flesh of wounded animals which crept to the streams to assuage their thirst, are those found in the cave earth. The absence of human bones may merely show that the mighty hunters of those days were too hardy, athletic, and intelligent, often to perish from accidental causes, and that they did not use this cavern for a place of burial. But the land again subsided. The valley of that now nameless river, of which the Rhine the Thames, and the Severn may have alike been tributaries, disappeared under the sea; and some tribe, driven from the lower lands, took refuge in this cave, now again near the encroaching waves, and left there the remains of their last repasts ere they were driven farther inland or engulfed in the waters. For a time the cavern may have been wholly submerged, and the charcoal of the extinguished fires became covered with its thin coating of clay. But ere long it re-emerged to form part of an island, long barren and desolate; and the valleys having been cut deeper by the receding waters, it no longer received muddy deposits, and the crust formed by drippings from its roof contained only bones and pebbles washed by rains or occasional land floods from its own clay deposits. Finally, the modern forests overspread the land, and were tenanted by the modern animals. Man returned to use the cavern again as a place of refuge or habitation, and to leave there the relics contained in the black earth. This seems at present the only intelligible history of this curious cave and others resembling it; though, when we consider the imperfection of the results obtained even by a large amount of labour, and the
difficult and confused character of the deposits in this and similar caves, too much value should not be attached to such histories, which may at any time be contradicted or modified by new facts or different explanations of those already known. The time involved depends very much, as already stated, on the question whether we regard the Post- glacial subsidence and re-elevation as somewhat sudden, or as occupying long ages at the slow rate at which some parts of our continents are now rising or sinking.[AT] [AT] Another element in this is also the question raised by Dawkins, Geikie, and others as to subdivisions of the Post-glacial period and intermissions of the Glacial cold. After careful consideration of these views, however, I cannot consider them as of much importance. Such are the glimpses, obscure though stimulating to the imagination, which geology can give of the circumstances attending the appearance of man in Western Europe. How far we are from being able to account for his origin, or to give its circumstances and relative dates for the whole world, the reader will readily understand. Still it is something to know that there is an intelligible meeting-place of the later geological ages and the age of man, and that it is one inviting to many and hopeful researches. It is curious also to find that the few monuments disinterred by geology, the antediluvian record of Holy Scripture, and the golden age of heathen tradition, seem alike to point to similar physical conditions, and to that simple state of the arts of life in which “gold and wampum and flint stones”[AU] constituted the chief material treasures of the earliest tribes of men. They also point to the immeasurable elevation, then as now, of man over his brute rivals for the dominion of the earth. To the naturalist this subject opens up most inviting yet most difficult paths of research, to be entered on with caution and reverence, rather than in the bold and dashing spirit of many modern attempts. The Christian, on his part, may feel satisfied that the scattered monumental relics of the caves and gravels will tell no story very different from that which he has long believed on other evidence, nor anything inconsistent with those views of man’s heavenly origin and destiny which have been the most precious inheritance of the greatest and best minds of every age, from that early pre-historic
period when men, “palaeolithic” men, no doubt, began to “invoke the name of Jehovah,” the coming Saviour, down to those times when life and immortality are brought to light, for all who will see, by the Saviour already come. [AU] So I read the “gold, bedolah, and shoham” of the description of Eden in Genesis ii.—the oldest literary record of the stone age. In completing this series of pictures, I wish emphatically to insist on the imperfection of the sketches which I have been able to present, and which are less, in comparison with the grand march of the creative work, even as now imperfectly known to science, than the roughest pencilling of a child when compared with a finished picture. If they have any popular value, it will be in presenting such a broad general view of a great subject as may induce further study to fill up the details. If they have any scientific value, it will be in removing the minds of British students for a little from the too exclusive study of their own limited marginal area, which has been to them too much the “celestial empire” around which all other countries must be arranged, and in divesting the subject of the special colouring given to it by certain prominent cliques and parties. Geology as a science is at present in a peculiar and somewhat exceptional state. Under the influence of a few men of commanding genius belonging to the generation now passing away, it has made so gigantic conquests that its armies have broken up into bands of specialists, little better than scientific banditti, liable to be beaten in detail, and prone to commit outrages on common sense and good taste, which bring their otherwise good cause into disrepute. The leaders of these bands are, many of them, good soldiers, but few of them fitted to be general officers, and none of them able to reunite our scattered detachments. We need larger minds, of broader culture and wider sympathies, to organise and rule the lands which we have subdued, and to lead on to further conquests. In the present state of natural science in Britain, this evil is perhaps to be remedied only by providing a wider and deeper culture for our young men. Few of our present workers have enjoyed that thorough
training in mental as well as physical science, which is necessary to enable men even of great powers to take large and lofty views of the scheme of nature. Hence we often find men who are fair workers in limited departments, reasoning most illogically, taking narrow and local views, elevating the exception into the rule, led away by baseless metaphysical subtleties, quarrelling with men who look at their specialties from a different point of view, and even striving and plotting for the advancement of their own hobbies. Such defects certainly mar much of the scientific work now being done. In the more advanced walks of scientific research, they are to some extent neutralised by that free discussion which true science always fosters; though even here they sometimes vexatiously arrest the progress of truth, or open floodgates of error which it may require much labour to close. But in public lectures and popular publications they run riot, and are stimulated by the mistaken opposition of narrow-minded good men, by the love of the new and sensational, and by the rivalry of men struggling for place and position. To launch a clever and startling fallacy which will float for a week and stir up a hard fight, seems almost as great a triumph as the discovery of an important fact or law; and the honest student is distracted with the multitude of doctrines, and hustled aside by the crowd of ambitious groundlings. The only remedy in the case is a higher and more general scientific education; and yet I do not wonder that many good men object to this, simply because of the difficulty of finding honest and competent teachers, themselves well grounded in their subjects, and free from that too common insanity of specialists and half-educated men, which impels them to run amuck at everything that does not depend on their own methods of research. This is a difficulty which can be met in our time only by the general good sense and right feeling of the community taking a firm hold of the matter, and insisting on the organization and extension of the higher scientific education, as well as that of a more elementary character, under the management of able and sane men. Yet even if not so counteracted, present follies will pass away, and a new and better state of natural science will arise in the future, by its own internal development. Science cannot long successfully isolate itself from God. Its life lies in the fact that it is the
exponent of the plans and works of the great Creative Will. It must, in spite of itself, serve His purposes, by dispelling blighting ignorance and superstition, by lighting the way to successive triumphs of human skill over the powers of nature, and by guarding men from the evils that flow from infringement of natural laws. And it cannot fail, as it approaches nearer to the boundaries of that which may be known by finite minds, to be humbled by the contemplation of the infinite, and to recognise therein that intelligence of which the human mind is but the image and shadow. It may be that theologians also are needed who shall be fit to take the place of Moses to our generation, in teaching it again the very elements of natural theology; but let them not look upon science as a cold and godless demon, holding forth to the world a poisoned cup cunningly compounded of truth and falsehood; but rather as the natural ally and associate of the gospel of salvation. The matter is so put in one of those visions which close the canon of revelation, when the prophet sees a mighty angel having the “everlasting gospel to preach;” but he begins his proclamation by calling on men to “worship Him that made heaven and earth and the sea and the fountains of waters.” Men must know God as the Creator even before they seek Him as a benefactor and redeemer. Thus religion must go hand in hand with all true and honest science. In this way only may we look forward to a time when a more exact and large-minded science shall be in perfect accord with a more pure and spiritual Christianity, when the natural and the spiritual shall be seen to be the necessary complements of each other, and when we shall hear no more of reconciliations between science and theology, because there will be no quarrels to reconcile. Already, even in the present chaos of scientific and religious opinion, indications can be seen by the observant, that the Divine Spirit of order is breathing on the mass, and will evolve from it new and beautiful worlds of mental and spiritual existence. CHAPTER XIV.
PRIMITIVE MAN. CONSIDERED WITH REFERENCE TO MODERN THEORIES AS TO HIS ORIGIN. The geological record, as we have been reading it, introduces us to primitive man, but gives us no distinct information as to his origin. Tradition and revelation have, it is true, their solutions of the mystery, but there are, and always have been, many who will not take these on trust, but must grope for themselves with the taper of science or philosophy into the dark caverns whence issue the springs of humanity. In former times it was philosophic speculation alone which lent its dim and uncertain light to these bold inquirers; but in our day the new and startling discoveries in physics, chemistry, and biology have flashed up with an unexpected brilliancy, and have at least served to dazzle the eyes and encourage the hopes of the curious, and to lead to explorations more bold and systematic than any previously undertaken. Thus has been born amongst us, or rather renewed, for it is a very old thing, that evolutionist philosophy, which has been well characterised as the “baldest of all the philosophies which have sprung up in our world,” and which solves the question of human origin by the assumption that human nature exists potentially in mere inorganic matter, and that a chain of spontaneous derivation connects incandescent molecules or star-dust with the world, and with man himself. This evolutionist doctrine is itself one of the strangest phenomena of humanity. It existed, and most naturally, in the oldest philosophy and poetry, in connection with the crudest and most uncritical, attempts of the human mind to grasp the system of nature; but that in our day a system destitute of any shadow of proof, and supported merely by vague analogies and figures of speech, and by the arbitrary and artificial coherence of its own parts, should be accepted as a philosophy, and should find able adherents to string upon its thread of hypotheses our vast and weighty stores of knowledge, is surpassingly strange. It seems to indicate that the accumulated facts of our age
have gone altogether beyond its capacity for generalisation; and but for the vigour which one sees everywhere, it might be taken as an indication that the human mind has fallen into a state of senility, and in its dotage mistakes for science the imaginations which were the dreams of its youth. In many respects these speculations are important and worthy of the attention of thinking men. They seek to revolutionise the religious beliefs of the world, and if accepted would destroy most of the existing theology and philosophy. They indicate tendencies among scientific thinkers, which, though probably temporary, must, before they disappear, descend to lower strata, and reproduce themselves in grosser forms, and with most serious effects on the whole structure of society. With one class of minds they constitute a sort of religion, which so far satisfies the craving for truths higher than those which relate to immediate wants and pleasures. With another and perhaps larger class, they are accepted as affording a welcome deliverance from all scruples of conscience and fears of a hereafter. In the domain of science evolutionism has like tendencies. It reduces the position of man, who becomes a descendant of inferior animals, and a mere term in a series whose end is unknown. It removes from the study of nature the ideas of final cause and purpose; and the evolutionist, instead of regarding the world as a work of consummate plan, skill, and adjustment, approaches nature as he would a chaos of fallen rocks, which may present forms of castles and grotesque profiles of men and animals, but they are all fortuitous and without significance. It obliterates the fine perception of differences from the mind of the naturalist, and resolves all the complicated relations of living things into some simple idea of descent with modification. It thus destroys the possibility of a philosophical classification, reducing all things to a mere series, and leads to a rapid decay in systematic zoology and botany, which is already very manifest among the disciples of Spencer and Darwin in England. The effect of this will be, if it proceeds further, in a great degree to destroy the educational value and popular interest attaching to these sciences, and to throw them down at the feet of a system of debased metaphysics. As redeeming features in all this, are the careful study of varietal forms, and the inquiries as to the limits of
species, which have sprung from these discussions, and the harvest of which will be reaped by the true naturalists of the future. Thus these theories as to the origin of men and animals and plants are full of present significance, and may be studied with profit by all; and in no part of their applications more usefully than in that which relates to man. Let us then inquire,—1. What is implied in the idea of evolution as applied to man? 2. What is implied in the idea of creation? 3. How these several views accord with what we actually know as the result of scientific investigation? The first and second of these questions may well occupy the whole of this chapter, and we shall be able merely to glance at their leading aspects. In doing so, it may be well first to place before us in general terms the several alternatives which evolutionists offer, as to the mode in which the honour of an origin from apes or ape-like animals can be granted to us, along with the opposite view as to the independent origin of man which have been maintained either on scientific or scriptural grounds. All the evolutionist theories of the origin of man depend primarily on the possibility of his having been produced from some of the animals more closely allied to him, by the causes now in operation which lead to varietal forms, or by similar causes which have been in operation; and some attach more and others less weight to certain of these causes, or gratuitously suppose others not actually known. Of such causes of change some are internal and others external to the organism. With respect to the former, one school assumes an innate tendency in every species to change in the course of time.[AV] Another believes in exceptional births, either in the course of ordinary generation or by the mode of parthenogenesis.[AW] Another refers to the known facts of reproductive accelleration or retardation observed in some humble creatures.[AX] New forms arising in any of these ways or fortuitously, may, it is supposed, be perpetuated and increased and further improved by favouring external circumstances and the effort of the organism to avail itself of these,[AY] or by the struggle for existence and the survival of the fittest.[AZ] [AV] Parsons, Owen.
[AW] Mivart, Ferris. [AX] Hyatt and Cope. [AY] Lamarck, etc. [AZ] Darwin, etc. On the other hand, those who believe in the independent origin of man admit the above causes as adequate only to produce mere varieties, liable to return into the original stock. They may either hold that man has appeared as a product of special and miraculous creation, or that he has been created mediately by the operation of forces also concerned in the production of other animals, but the precise nature of which is still unknown to us; or lastly, they may hold what seems to be the view favoured by the book of Genesis, that his bodily form is a product of mediate creation and his spiritual nature a direct emanation from his Creator. The discussion of all these rival theories would occupy volumes, and to follow them into details would require investigations which have already bewildered many minds of some scientific culture. Further, it is the belief of the writer that this plunging into multitudes of details has been fruitful of error, and that it will be a better course to endeavour to reach the root of the matter by looking at the foundations of the general doctrine of evolution itself, and then contrasting it with its rival. Taking, then, this broad view of the subject, two great leading alternatives are presented to us. Either man is an independent product of the will of a Higher Intelligence, acting directly or through the laws and materials of his own institution and production, or he has been produced by an unconscious evolution from lower things. It is true that many evolutionists, either unwilling to offend, or not perceiving the logical consequences of their own hypothesis, endeavour to steer a middle course, and to maintain that the Creator has proceeded by way of evolution. But the bare, hard logic of Spencer, the greatest English authority on evolution, leaves no place for this compromise, and shows that the theory, carried out to its legitimate consequences, excludes the knowledge of a Creator and the possibility of His work. We have, therefore, to choose between evolution and creation; bearing in mind,
however, that there may be a place in nature for evolution, properly limited, as well as for other things, and that the idea of creation by no means excludes law and second causes. Limiting ourselves in the first place to theories of evolution, and to these as explaining the origin of species of living beings, and especially of man, we naturally first inquire as to the basis on which they are founded. Now no one pretends that they rest on facts actually observed, for no one has ever observed the production of even one species. Nor do they even rest, like the deductions of theoretical geology, on the extension into past time of causes of change now seen to be in action. Their probability depends entirely on their capacity to account hypothetically for certain relations of living creatures to each other, and to the world without; and the strongest point of the arguments of their advocates is the accumulation of cases of such relations supposed to be accounted for. Such being the kind of argument with which we have to deal, we may first inquire what we are required to believe as conditions of the action of evolution, and secondly, to what extent it actually does explain the phenomena. In the first place, as evolutionists, we are required to assume certain forces, or materials, or both, with which evolution shall begin. Darwin, in his Origin of Species, went so far as to assume the existence of a few of the simpler types of animals; but this view, of course, was only a temporary resting-place for his theory. Others assume a primitive protoplasm, or physical basis of life, and arbitrarily assigning to this substance properties now divided between organised and unorganised, and between dead and living matter, find no difficulty in deducing all plants and animals from it. Still, even this cannot have been the ultimate material. It must have been evolved from something. We are thus brought back to certain molecules of star-dust, or certain conflicting forces, which must have had self-existence, and must have potentially included all subsequent creatures. Otherwise, if with Spencer we hold that God is “unknowable” and creation “unthinkable,” we are left suspended on nothing over a bottomless void, and must adopt as the initial proposition of our philosophy, that all things were made out of nothing, and by nothing; unless we prefer
to doubt whether anything exists, and to push the doctrine of relativity to the unscientific extreme of believing that we can study the relations of things non-existent or unknown. So we must allow the evolutionist some small capital to start with; observing, however, that self-existent matter in a state of endless evolution is something of which we cannot possibly have any definite conception. Being granted thus much, the evolutionist next proceeds to demand that we shall also believe in the indefinite variability of material things, and shall set aside all idea that there is any difference in kind between the different substances which we know. They must all be mutually convertible, or at least derivable from some primitive material. It is true that this is contrary to experience. The chemist holds that matter is of different kinds, that one element cannot be converted into another; and he would probably smile if told that, even in the lapse of enormous periods of time, limestone could be evolved out of silica. He may think that this is very different from the idea that a snail can be evolved from an oyster, or a bird from a reptile. But the zoologist will inform him that species of animals are only variable within certain limits, and are not transmutable, in so far as experience and experiment are concerned. They have their allotropic forms, but cannot be changed into one another. But if we grant this second demand, the evolutionist has a third in store for us. We must also admit that by some inevitable necessity the changes of things must in the main take place in one direction, from the more simple to the more complex, from the lower to the higher. At first sight this seems not only to follow from the previous assumptions, but to accord with observation. Do not all living things rise from a simpler to a more complex state? has not the history of the earth displayed a gradually increasing elevation and complexity? But, on the other hand, the complex organism becoming mature, resolves itself again into the simple germ, and finally is dissolved into its constituent elements. The complex returns into the simple, and what we see is not an evolution, but a revolution. In like manner, in geological time, the tendency seems to be ever to disintegration and decay. This we see everywhere, and find that elevation occurs only by the introduction of
new species in a way which is not obvious, and which may rather imply the intervention of a cause from without; so that here also we are required to admit as a general principle what is contrary to experience. If, however, we grant the evolutionist these postulates, we must next allow him to take the facts of botany and zoology out of their ordinary connection, and thread them like a string of beads, as Herbert Spencer has done in his “Biology,” on the threefold cord thus fashioned. This done, we next find, as might have been expected, certain gaps or breaks which require to be cunningly filled with artificial material, in order to give an appearance of continuity to the whole. The first of these gaps which we notice is that between dead and living matter. It is easy to fill this with such a term as protoplasm, which includes matter both dead and living, and so to ignore this distinction; but practically we do not yet know as a possible thing the elevation of matter, without the agency of a previous living organism, from that plane in which it is subject merely to physical force, and is unorganised, to that where it becomes organised, and lives. Under that strange hypothesis of the origin of life from meteors, with which Sir William Thomson closed his address at a late meeting of the British Association, there was concealed a cutting sarcasm which the evolutionists felt. It reminded them that the men who evolve all things from physical forces do not yet know how these forces can produce the phenomena of life even in its humblest forms. It is true that the scientific world has been again and again startled by the announcement of the production of some of the lowest forms of life, either from dead organic matter, or from merely mineral substances; but in every case heretofore the effort has proved as vain as the analogies attempted to be set up between the formation of crystals and that of organized tissues are fallacious. A second gap is that which separates vegetable and animal life. These are necessarily the converse of each other, the one deoxidizes and accumulates, the other oxidizes and expends. Only in reproduction or decay does the plant simulate the action of the animal, and the animal never in its simplest forms assumes the functions of the plant. Those obscure cases in the humbler spheres of animal and vegetable
life which have been supposed to show a union of the two kingdoms, disappear on investigation. This gap can, I believe, be filled up only by an appeal to our ignorance. There may be, or may have been, some simple creature unknown to us, on the extreme verge of the plant kingdom, that was capable of passing the limit and becoming an animal. But no proof of this exists. It is true that the primitive germs of many kinds of humble plants and animal s are so much alike, that much confusion has arisen in tracing their development. It is also true that some of these creatures can subsist under very dissimilar conditions, and in very diverse states, and that under the specious name of Biology,[BA] we sometimes find a mass of these confusions, inaccurate observations and varietal differences made to do duty for scientific facts. But all this does not invalidate the grand primary distinction between the animal and the plant, which should be thoroughly taught and illustrated to all young naturalists, as one of the best antidotes to the fallacies of the evolutionist school. [BA] It is doubtful whether men who deny the existence of vital force have a right to call their science “Biology,” any more than atheists have to call their doctrine “Theology;” and it is certain that the assumption of a science of Biology as distinct from Phytology and Zoology, or including both, is of the nature of a “pious fraud” on the part of the more enlightened evolutionists. The objections stated in the text, to what have been called Archebiosis and Heterogenesis seem perfectly applicable, in so far as I can judge from a friendly review by Wallace, to the mass of heterogeneous material accumulated by Dr. Bastian in his recent volumes. The conclusions of this writer, would also, if established, involve evolution in a fatal embarras des richesses, by the hourly production during all geological time, of millions of new forms all capable of indefinite development. A third is that between any species of animal or plant and any other species. It was this gap, and this only, which Darwin undertook to fill up by his great work on the origin of species, but, notwithstanding the immense amount of material thus expended, it yawns as wide as ever, since it must be admitted that no case has been ascertained in which an individual of one species has transgressed the limits between it and other species. However extensive the varieties produced by artificial breeding, the essential characters of the species remain, and even its
minor characters may be reproduced, while the barriers established in nature between species by the laws of their reproduction, seem to be absolute. With regard to species, however, it must be observed that naturalists are not agreed as to what constitutes a species. Many so- called species are probably races, or varieties, and one benefit of these inquiries has been to direct attention to the proper discrimination of species from varieties among animals and plants. The loose discrimination of species, and the tendency to multiply names, have done much to promote evolutionist views; but the researches of the evolutionists themselves have shown that we must abandon transmutation of true species as a thing of the present; and if we imagine it to have occurred, must refer it to the past. Another gap is that between the nature of the animal and the self- conscious, reasoning, moral nature of man. We not only have no proof that any animal can, by any force in itself, or by any merely physical influences from without, rise to such a condition; but the thing is in the highest degree improbable. It is easy to affirm, with the grosser materialists, that thought is a secretion of brain, as bile is of the liver; but a moment’s thought shows that no real analogy obtains between the cases. We may vaguely suppose, with Darwin, that the continual exercise of such powers as animals possess, may have developed those of man. But our experience of animals shows that their intelligence differs essentially from that of man, being a closed circle ever returning into itself, while that of man is progressive, inventive, and accumulative, and can no more be correlated with that of the animal than the vital phenomena of the animal with those of the plant. Nor can the gap between the higher religious and moral sentiments of man, and the instinctive affections of the brutes, be filled up with that miserable ape imagined by Lubbock, which, crossed in love, or pining with cold and hunger, conceived, for the first time in its poor addled pate, “the dread of evil to come,” and so became the father of theology. This conception, which Darwin gravely adopts, would be most ludicrous, but for the frightful picture which it gives of the aspect in which religion appears to the mind of the evolutionist.
The reader will now readily perceive that the simplicity and completeness of the evolutionist theory entirely disappear when we consider the unproved assumptions on which it is based, and its failure to connect with each other some of the most important facts in nature: that, in short, it is not in any true sense a philosophy, but merely an arbitrary arrangement of facts in accordance with a number of unproved hypotheses. Such philosophies, “falsely so called,” have existed ever since man began to reason on nature, and this last of them is one of the weakest and most pernicious of the whole. Let the reader take up either of Darwin’s great books, or Spencer’s “Biology,” and merely ask himself as he reads each paragraph, “What is assumed here and what is proved?” and he will find the whole fabric melt away like a vision. He will find, however, one difference between these writers. Darwin always states facts carefully and accurately, and when he comes to a difficulty tries to meet it fairly. Spencer often exaggerates or extenuates with reference to his facts, and uses the arts of the dialectician where argument fails. Many naturalists who should know better are puzzled with the great array of facts presented by evolutionists; and while their better judgment causes them to doubt as to the possibility of the structures which they study being produced by such blind and material processes, are forced to admit that there must surely be something in a theory so confidently asserted, supported by so great names, and by such an imposing array of relations which it can explain. They would be relieved from their weak concessions were they to study carefully a few of the instances adduced, and to consider how easy it is by a little ingenuity to group undoubted facts around a false theory. I could wish to present here illustrations of this, which abound in every part of the works I have referred to, but space will not permit. One or two must suffice. The first may be taken from one of the strong points often dwelt on by Spencer in his “Biology.”[BB] [BB] “Principles of Biology,” § 118. But the experiences which most clearly illustrate to us the process of general evolution are our experiences of special evolution, repeated in every plant and animal. Each organism exhibits, within a short space
of time, a series of changes which, when supposed to occupy a period indefinitely great and to go on in various ways instead of one, may give us a tolerably clear conception of organic evolution in general. In an individual development we have compressed into a comparatively infinitesimal space a series of metamorphoses equally vast with those which the hypothesis of evolution assumes to have taken place during those unmeasurable epochs that the earth’s crust tells us of. A tree differs from a seed immeasurably in every respect—in bulk, in structure, in colour, in form, in specific gravity, in chemical composition: differs so greatly that no visible resemblance of any kind can be pointed out between them. Yet is the one changed in the course of a few years into the other; changed so gradually that at no moment can it be said, ‘Now the seed ceases to be and the tree exists.’ What can be more widely contrasted than a newly-born child and the small gelatinous spherule constituting the human ovum? The infant is so complex in structure that a cyclopædia is needed to describe its constituent parts. The germinal vesicle is so simple that it may be defined in a line.... If a single cell under appropriate conditions becomes a man in the space of a few years, there can surely be no difficulty in understanding how, under appropriate conditions, a cell may in the course of untold millions of years give origin to the human race. “It is true that many minds are so unfurnished with those experiences of nature, out of which this conception is built, that they find difficulty in forming it.... To such the hypothesis that by any series of changes a protozoan should ever give origin to a mammal seems grotesque—as grotesque as did Galileo’s assertion of the earth’s movement seem to the Aristoteleans; or as grotesque as the assertion of the earth’s sphericity seems now to the New Zealanders.” I quote the above as a specimen of evolutionist reasoning from the hand of a master, and as referring to one of the corner-stones of this strange philosophy. I may remark with respect to it, in the first place, that it assumes those “conditions” of evolution to which I have already referred. In the second place, it is full of inaccurate statements of fact, all in a direction tending to favour the hypothesis. For example, a tree
does not differ “immeasurably” from a seed, especially if the seed is of the same species of tree, for the principal parts of the tree and its principal chemical constituents already exist and can be detected in the seed, and unless it were so, the development of the tree from the seed could not take place. Besides, the seed itself is not a thing self-existent or fortuitous. The production of a seed without a previous tree of the same kind is quite as difficult to suppose as the production of a tree without a previous seed containing its living embryo. In the third place, the whole argument is one of analogy. The germ becomes a mature animal, passing through many intermediate stages, therefore the animal may have descended from some creature which when mature was as simple as the germ. The value of such an analogy depends altogether on the similarity of the “conditions” which, in such a case, are really the efficient causes at work. The germ of a mammal becomes developed by the nourishment supplied from the system of a parent, which itself produced the germ, and into whose likeness the young animal is destined to grow. These are the “appropriate conditions” of its development. But when our author assumes from this other “appropriate conditions,” by which an organism, which on the hypothesis is not a germ but a mature animal, shall be developed into the likeness, of something different from its parent, he oversteps the bounds of legitimate analogy. Further, the reproduction of the animal, as observed, is a closed series, beginning at the embryo and returning thither again; the evolution attempted to be established is a progressive series going on from one stage to another. A reproductive circle once established obeys certain definite laws, but its origin, or how it can leave its orbit and revolve in some other, we cannot explain without the introduction of some new efficient cause. The one term of the analogy is a revolution, and the other is an evolution. The revolution within the circle of the reproduction of the species gives no evidence that at some point the body will fly off at a tangent, and does not even inform us whether it is making progress in space. Even if it is so making progress, its orbit of revolution may remain the same. But it may be said the reproduction of the species is not in a circle but in a spiral. Within the limit of experience it is not so, since, however it may undulate, it always returns into itself. But supposing it to be a spiral, it may ascend or descend, or expand and contract; but this does not
connect it with other similar spirals, the separate origin of which is to be separately accounted for. I have quoted the latter part of the passage because it is characteristic of evolutionists to decry the intelligence of those who differ from them. Now it is fair to admit that it requires some intelligence and some knowledge of nature to produce or even to understand such analogies as those of Mr. Spencer and his followers, but it is no less true that a deeper insight into the study of nature may not only enable us to understand these analogies, but to detect their fallacies. I am sorry to say, however, that at present the hypothesis of evolution is giving so strong a colouring to much of popular and even academic teaching, more especially in the easy and flippant conversion of the facts of embryology into instances of evolution on the plan of the above extract, that the Spencerians may not long have to complain of want of faith and appreciation on the part of the improved apes whom they are kind enough to instruct as to their lowly origin. The mention of “appropriate conditions” in the above extract reminds me of another fatal objection to evolution which its advocates continually overlook. An animal or plant advancing from maturity to the adult state is in every stage of its progress a complete and symmetrical organism, correlated in all its parts and adapted to surrounding conditions. Suppose it to become modified in any way, to ever so small an extent, the whole of these relations are disturbed. If the modification is internal and spontaneous, there is no guarantee that it will suit the vastly numerous external agencies to which the creature is subjected. If it is produced by agencies from without, there is no guarantee that it will accord with the internal relations of the parts modified. The probabilities are incalculably great against the occurrence of many such disturbances without the breaking up altogether of the nice adjustment of parts and conditions. This is no doubt one reason of the extinction of so many species in geological time, and also of the strong tendency of every species to spring back to its normal condition when in any way artificially caused to vary. It is also connected with the otherwise mysterious law of the constant transmission of all the characters of the parent.
Spencer and Darwin occasionally see this difficulty, though they habitually neglect it in their reasonings. Spencer even tries to turn one part of it to account as follows:— “Suppose the head of a mammal to become very much more weighty—what must be the indirect results? The muscles of the neck are put to greater exertions; and the vertebras have to bear additional tensions and pressures caused both by the increased weight of the head and the stronger contraction of muscles that support and move the head.” He goes on to say that the processes of the vertebrae will have augmented strains put upon them, the thoracic region and fore limbs will have to be enlarged, and even the hind limbs may require modification to facilitate locomotion. He concludes: “Any one who compares the outline of the bison with that of its congener, the ox, will clearly see how profoundly a heavier head affects the entire osseous and muscular system.” We need not stop to mention the usual inaccuracies as to facts in this paragraph, as, for example, the support of the head being attributed to muscles alone, without reference to the strong elastic ligament of the neck. We may first notice the assumption that an animal can acquire a head “very much more weighty” than that which it had before, a very improbable supposition, whether as a monstrous birth Dr as an effect of external conditions after birth. But suppose this to have occurred, and what is even less likely, that the very much heavier head is an advantage in some way, what guarantee can evolution give us that the number of other modifications required would take place simultaneously with this acquisition! It would be easy to show that this would depend on the concurrence of hundreds of other conditions within and without the animal, all of which must co- operate to produce the desired effect, if indeed they could produce this effect even by their conjoint action, a power which the writer, it will be observed, quietly assumes, as well as the probability of the initial change in the head. Finally, the naivete with which it is assumed that the bison and the ox are examples of such an evolution, would be refreshing in these artificial days, if instances of it did not occur in almost every page of the writings of evolutionists.
It would only weary the reader to follow evolution any further into details, especially as my object in this chapter is to show that generally, and as a theory of nature and of man, it has no good foundation; but we should not leave the subject without noting precisely the derivation of man according to this theory; and for this purpose I may quote Darwin’s summary of his conclusions on the subject.[BC] [BC] “Descent of Man,” part ii., ch. 21. “Man,” says Mr. Darwin, “is descended from a hairy quadruped, furnished with a tail and pointed ears, probably arboreal in its habits, and an inhabitant of the Old World. This creature, if its whole structure had been examined by a naturalist, would have been classed amongst the quadrumana, as surely as would the common, and still more ancient, progenitor of the Old and New World monkeys. The quadrumana and all the higher mammals are probably derived from an ancient marsupial animal; and this, through a long line of diversified forms, either from some reptile-like or some amphibian-like creature, and this again from some fish-like animal. In the dim obscurity of the past we can see that the early progenitor of all the vertebrata must have been an aquatic animal, provided with branchiæ, with the two sexes united in the same individual, and with the most important organs of the body (such as the brain and heart) imperfectly developed. This animal seems to have been more like the larvæ of our existing marine Ascidians than any other form known.” The author of this passage, in condescension to our weakness of faith, takes us no further back than to an Ascidian, or “sea-squirt,” the resemblance, however, of which to a vertebrate animal is merely analogical, and, though a very curious case of analogy, altogether temporary and belonging to the young state of the creature, without affecting its adult state or its real affinities with other mollusks. In order, however, to get the Ascidian itself, he must assume all the “conditions” already referred to in the previous part of this article, and fill most of the gaps. He has, however, in the “Origin of Species” and “Descent of Man,” attempted merely to fill one of the breaks in the evolutionary series, that between distinct species, leaving us to receive
all the rest on mere faith. Even in respect to the question of species, in all the long chain between the Ascidian and the man, he has not certainly established one link; and in the very last change, that from the ape-like ancestor, he equally fails to satisfy us as to matters so trivial as the loss of the hair, which, on the hypothesis, clothed the pre- human back, and on matters so weighty as the dawn of human reason and conscience. We thus see that evolution as an hypothesis has no basis in experience or in scientific fact, and that its imagined series of transmutations has breaks which cannot be filled. We have now to consider how it stands with the belief that man has been created by a higher power. Against this supposition the evolutionists try to create a prejudice in two ways. First, they maintain with Herbert Spencer that the hypothesis of creation is inconceivable, or, as they say, “unthinkable;” an assertion which, when examined, proves to mean only that we do not know perfectly the details of such an operation, an objection equally fatal to the origin either of matter or life, on the hypothesis of evolution. Secondly, they always refer to creation as if it must be a special miracle, in the sense of a contravention of or departure from ordinary natural laws; but this is an assumption utterly without proof, since creation may be as much according to law as evolution, though in either case the precise laws involved may be very imperfectly known. How absurd, they say, to imagine an animal created at once, fully formed, by a special miracle, instead of supposing it to be slowly elaborated through, countless ages of evolution. To Darwin the doctrine of creation is but “a curious illustration of the blindness of preconceived opinion.” “These authors,” he says, “seem no more startled at a miraculous act of creation than at an ordinary birth; but do they really believe that at innumerable periods in the earth’s history, certain elemental atoms have been commanded suddenly to flash into living tissues?” Darwin, with all his philosophic fairness, sometimes becomes almost Spencerian in his looseness of expression; and in the above extract, the terms “miraculous,” “innumerable,” “elemental atoms,” “suddenly,” and “flash,” all express ideas in no respect
necessary to the work of creation. Those who have no faith in evolution as a cause of the production of species, may well ask in return how the evolutionist can prove that creation must be instantaneous, that it must follow no law, that it must produce an animal fully formed, that it must be miraculous. In short, it is a portion of the policy of evolutionists to endeavour to tie down their opponents to a purely gratuitous and ignorant view of creation, and then to attack them in that position. What, then, is the actual statement of the theory of creation as it may be held by a modern man of science? Simply this; that all things have been produced by the Supreme Creative Will, acting either directly or through the agency of the forces and materials of His own production. This theory does not necessarily affirm that creation is miraculous, in the sense of being contrary to or subversive of law; law and order are as applicable to creation as to any other process. It does not contradict the idea of successive creations. There is no necessity that the process should be instantaneous and without progression. It does not imply that all kinds of creation are alike. There may be higher and lower kinds. It does not exclude the idea of similarity or dissimilarity of plan and function as to the products of creation. Distinct products of creation may be either similar to each other in different degrees, or dissimilar. It does not even exclude evolution or derivation to a certain extent: anything once created may, if sufficiently flexible and elastic, be evolved or involved in various ways. Indeed, creation and derivation may, rightly understood, be complementary to each other. Created things, unless absolutely unchangeable, must be more or less modified by influences from within and from without, and derivation or evolution may account for certain subordinate changes of things already made. Man, for example, may be a product of creation, yet his creation may have been in perfect harmony with those laws of procedure which the Creator has set for His own operations. He may have been preceded by other creations of things more or less similar or dissimilar. He may have been created by the same processes with some or all of these, or by different means. His body may have been created in one way, his soul
in another. He may, nay, in all probability would be, part of a plan of which some parts would approach very near to him in structure or functions. After his creation, spontaneous culture and outward circumstances may have moulded him into varieties, and given him many different kinds of speech and of habits. These points are so obvious to common sense that it would be quite unnecessary to insist on them, were they not habitually overlooked or misstated by evolutionists. The creation hypothesis is also free from some of the difficulties of evolution. It avoids the absurdity of an eternal progression from the less to the more complex. It provides in will, the only source of power actually known to us by ordinary experience, an intelligible origin of nature. It does not require us to contradict experience by supposing that there are no differences of kind or essence in things. It does not require us to assume, contrary to experience, an invariable tendency to differentiate and improve. It does not exact the bridging over of all gaps which may be found between the several grades of beings which exist or have existed. Why, then, are so many men of science disposed to ignore altogether this view of the matter? Mainly, I believe, because, from the training of many of them, they are absolutely ignorant of the subject, and from their habits of thought have come to regard physical force and the laws regulating it as the one power in nature, and to relegate all spiritual powers or forces, or, as they have been taught to regard them, “supernatural” things, to the domain of the “unknowable.” Perhaps some portion of the difficulty may be got over by abandoning altogether the word “supernatural,” which has been much misused, and by holding nature to represent the whole cosmos, and to include both the physical and the spiritual, both of them in the fullest sense subject to law, but each to the law of its own special nature. I have read somewhere a story of some ignorant orientals who were induced to keep a steam-engine supplied with water by the fiction that it contained a terrible djin, or demon, who, if allowed to become thirsty, would break out and destroy them all. Had they been enabled to discard this superstition, and to understand the force of steam, we can
readily imagine that they would now suppose they knew the whole truth, and might believe that any one who taught them that the engine was a product of intelligent design, was only taking them back to the old doctrine of the thirsty demon of the boiler. This is, I think, at present, the mental condition of many scientists with reference to creation. Here we come to the first demand which the doctrine of creation makes on us by way of premises. In order that there may be creation there must be a primary Self-existent Spirit, whose will is supreme. The evolutionist cannot refuse to admit this on as good ground as that on which we hesitate to receive the postulates of his faith. It is no real objection to say that a God can be known to us only partially, and, with reference to His real essence, not at all; since, even if we admit this, it is no more than can be said of matter and force. I am not about here to repeat any of the ordinary arguments for the existence of a spiritual First Cause, and Creator of all things, but it may be proper to show that this assumption is not inconsistent with experience, or with the facts and principles of modern science. The statement which I would make on this point shall be in the words of a very old writer, not so well known as he should be to many who talk volubly enough about antagonisms between science and Christianity: “That which is known of God is manifest in them (in men), for God manifested it unto them. For since the creation of the world His invisible things, even His eternal power and divinity are plainly seen, being perceived by means of things that are made.”[BD] The statement here is very precise. Certain things relating to God are manifest within men’s minds, and are proved by the evidence of His works; these properties of God thus manifested being specially His power or control of all forces, and His divinity or possession of a nature higher than ours. The argument of the writer is that all heathens know this; and, as a matter of fact, I believe it must be admitted even by those most sceptical on such points, that some notion of a divinity has been derived from nature by men of all nations and tribes, if we except, perhaps, a few enlightened positivists of this nineteenth century whom excess of light has made blind. “If the light that is in man be darkness,
how great is that darkness.” But then this notion of a God is a very old and primitive one, and Spencer takes care to inform us that “first thoughts are either wholly out of harmony with things, or in very incomplete harmony with them,” and consequently that old beliefs and generally diffused notions are presumably wrong. [BD] Paul’s Epistle to the Romans, chap i. Is it true, however, that the modern knowledge of nature tends to rob it of a spiritual First Cause? One can conceive such a tendency, if all our advances in knowledge had tended more and more to identify force with matter in its grosser forms, and to remove more and more from our mental view those powers which are not material; but the very reverse of this is the case. Modern discovery has tended more and more to attach importance to certain universally diffused media which do not seem to be subject to the laws of ordinary matter, and to prove at once the Protean character and indestructibility of forces, the aggregate of which, as acting in the universe, gives us our nearest approach to the conception of physical omnipotence. This is what so many of our evolutionists mean when they indignantly disclaim materialism. They know that there is a boundless energy beyond mere matter, and of which matter seems the sport and toy. Could they conceive of this energy as the expression of a personal will, they would become theists. Man himself presents a microcosm of matter and force, raised to a higher plane than that of the merely chemical and physical. In him we find not merely that brain and nerve force which is common to him and lower animals, and which exhibits one of the most marvellous energies in nature, but we have the higher force of will and intellect, enabling him to read the secrets of nature, to seize and combine and utilize its laws like a god, and like a god to attain to the higher discernment of good and evil. Nay, more, this power which resides within man rules with omnipotent energy the material organism, driving its nerve forces until cells and fibres are worn out and destroyed, taxing muscles and tendons till they break, impelling its slave the body even to that which will bring injury and death itself. Surely, what we thus see in man must be the image and likeness of the Great Spirit. We can escape from this
conclusion only by one or other of two assumptions, either of which is rather to be called a play upon words than a scientific theory. We may, with a certain class of physicists and physiologists, confine our attention wholly to the fire and the steam, and overlook the engineer. We may assume that with protoplasm and animal electricity, for example, we can dispense with life, and not only with life but with spirit also. Yet he who regards vitality as an unmeaning word; and yet speaks of “living protoplasm,” and “dead protoplasm,” and affirms that between these two states, so different in their phenomena, no chemical or physical difference exists, is surely either laughing at us, or committing himself to what the Duke of Argyll calls a philosophical bull; and he who shows us that electrical discharges are concerned in muscular contraction, has just as much proved that there is no need of life or spirit, as the electrician who has explained the mysteries of the telegraph has shown that there can be no need of an operator. Or we may, turning to the opposite extreme, trust to the metaphysical fallacy of those who affirm that neither matter, nor force, nor spirit, need concern them, for that all are merely states of consciousness in ourselves. But what of the conscious self this self which thinks, and which is in relation with surroundings which it did not create, and which presumably did not create it? and what is the unknown third term which must have been the means of setting up these relations? Here again our blind guides involve us in an absolute self-contradiction. Thus we are thrown back on the grand old truth that man, heathen and savage, or Christian and scientific, opens his eyes on nature and reads therein both the physical and the spiritual, and in connection with both of these the power and divinity of an Almighty Creator. He may at first have many wrong views both of God and of His works, but as he penetrates further into the laws of matter and mind, he attains more just conceptions of their relations to the Great Centre and Source of all, and instead of being able to dispense with creation, he hopes to be able at length to understand its laws and methods. If unhappily he abandons this high ambition, and contents himself with mere matter and physical force, he cannot rise to the highest development either of science or philosophy.
It may, however, be said that evolution may admit all this, and still be held as a scientific doctrine in connection with a modified belief in creation. The work of actual creation may have been limited to a few elementary types, and evolution may have done the rest. Evolutionists may still be theists. We have already seen that the doctrine, as carried out to its logical consequences, excludes creation and theism. It may, however, be shown that even in its more modified forms, and when held by men who maintain that they are not atheists, it is practically atheistic, because excluding the idea of plan and design, and resolving all things into the action of unintelligent forces. It is necessary to observe this, because it is the half-way evolutionism which professes to have a Creator somewhere behind it, that is most popular; though it is, if possible, more unphilosophical than that which professes to set out from absolute and eternal nonentity, or from self-existent star-dust containing all the possibilities of the universe. Absolute atheists recognise in Darwinism, for example, a philosophy which reduces all things to a “gradual summation of innumerable minute and accidental material operations,” and in this they are more logical than those who seek to reconcile evolution with design. Huxley, in his “lay sermons,” referring to Paley’s argument for design founded on the structure of a watch, says that if the watch could be conceived to be a product of a less perfect structure improved by natural selection, it would then appear to be the “result of a method of trial and error worked by unintelligent agents, as likely as of the direct application of the means appropriate to that end, by an intelligent agent.” This is a bold and true assertion of the actual relation of even this modified evolution to rational and practical theism, which requires not merely this God “afar off,” who has set the stone of nature rolling and then turned His back upon it, but a present God, whose will is the law of nature, now as in times past. The evolutionist is really in a position of absolute antagonism to the idea of creation, even when held with all due allowance for the variations of created things within certain limits. Perhaps Paley’s old illustration of the watch, as applied by Huxley, may serve to show this as well as any other. If the imperfect watch,
useless as a time-keeper, is the work of the contriver, and the perfection of it is the result of unintelligent agents working fortuitously, then it is clear that creation and design have a small and evanescent share in the construction of the fabric of nature. But is it really so? Can we attribute the perfection of the watch to “accidental material operations” any more than the first effort to produce such an instrument? Paley himself long ago met this view of the case, but his argument may be extended by the admissions and pleas of the evolutionists themselves. For example, the watch is altogether a mechanical thing, and this fact by no means implies that it could not be made by an intelligent and spiritual designer, yet this assumption that physical laws exclude creation and design turns up in almost every page of the evolutionists. Paley has well shown that if the watch contained within itself machinery for making other watches, this would not militate against his argument. It would be so if it could be proved that a piece of metal had spontaneously produced an imperfect watch, and this a more perfect one, and so on; but this is precisely what evolutionists still require to prove with respect both to the watch and to man. On the other hand it is no argument for the evolution of the watch that there may be different kinds of watches, some more and others less perfect, and that ruder forms may have preceded the more perfect. This is perfectly compatible with creation and design. Evolutionists, however, generally fail to make this distinction. Nor would it be any proof of the evolution of the watch to find that, as Spencer would say, it was in perfect harmony with its environment, as, for instance, that it kept time with the revolution of the earth, and contained contrivances to regulate its motion under different temperatures, unless it could be shown that the earth’s motion and the changes of temperature had been efficient causes of the motion and the adjustments of the watch; otherwise the argument would look altogether in the direction of design. Nor would it be fair to shut up the argument of design to the idea that the watch must have suddenly flashed into existence fully formed and in motion. It would be quite as much a creation if slowly and laboriously made by the hand of the artificer, or if more rapidly struck off by machinery; and if the latter, it would not follow that the machine which produced the watch was at all like the watch itself. It might have been something very different.
Finally, when Spencer tries to cut at the root of the whole of this argument, by affirming that man has no more right to reason from himself with regard to his Maker than a watch would have to reason from its own mechanical structure and affirm the like of its maker, he signally fails. If the watch had such power of reasoning, it would be more than mechanical, and would be intelligent like its maker; and in any case, if thus reasoning it came to the conclusion that it was a result of “accidental material operations,” it would be altogether mistaken. Nor would it be nearer the truth if it held that it was a product of spontaneous evolution from an imperfect and comparatively useless watch that had been made millions of years before. We have taken this illustration of the watch merely as given to us by Huxley, and without in the least seeking to overlook the distinction between a dead machine and a living organism; but the argument for creation and design is quite as strong in the case of the latter, so long as it cannot be proved by actual facts to be a product of derivation from a distinct species. This has not been proved either in the care of man or any other species; and so long as it has not, the theory of creation and design is infinitely more rational and scientific than that of evolution in any of its forms. But all this does not relieve us from the question, How can species be created?—the same question put to Paul by the sceptics of the first century with reference to the resurrection—“How are the dead raised, and with what bodies do they come?” I do not wish to evade this question, whether applied to man or to a microscopic animalcule, and I would answer it with the following statements:— 1. The advocate of creation is in this matter in no worse position than the evolutionist. This we have already shown, and I may refer here to the fact that Darwin himself assumes at least one primitive form of animal and plant life, and he is confessedly just as little able to imagine this one act of creation as any other that may be demanded of him. 2. We are not bound to believe that all groups of individual animals, which naturalists may call species, have been separate products of
creation. Man himself has by some naturalists been divided into several species; but we may well be content to believe the creation of one primitive form, and the production of existing races by variation. Every zoologist and botanist who has studied any group of animals or plants with care, knows that there are numerous related forms passing into each other, which some naturalists might consider to be distinct species, but which it is certainly not necessary to regard as distinct products of creation. Every species is more or less variable, and this variability may be developed by different causes. Individuals exposed to unfavourable conditions will be stunted and depauperated; those in more favourable circumstances may be improved and enlarged. Important changes may thus take place without transgressing the limits of the species, or preventing a return to its typical forms; and the practice of confounding these more limited changes with the wider structural and physiological differences which separate true species is much to be deprecated. Animals which pass through metamorphoses, or which, are developed through the instrumentality of intermediate forms or “nurses”[BE] are not only liable to be separated by mistake into distinct species, but they may, tinder certain circumstances, attain to a premature maturity, or may be fixed for a time or permanently in an immature condition. Further, species, like individuals, probably have their infancy, maturity, and decay in geological time, and may present differences in these several stages. It is the remainder of true specific types left after all these sources of error are removed, that creation has to account for; and to arrive at this remainder, and to ascertain its nature and amount, will require a vast expenditure of skilful and conscientious labour. [BE] Mr. Mungo Ponton, in his book “The Beginning,” has based a theory of derivation on this peculiarity. 3. Since animals and plants have been introduced upon our earth in long succession throughout geologic time, and this in a somewhat regular manner, we have a right to assume that their introduction has been in accordance with a law or plan of creation, and that this may have included the co-operation of many efficient causes, and may have differed in its application to different cases. This is a very old doctrine of theology, for it appears in the early chapters of Genesis. There the
Welcome to our website – the perfect destination for book lovers and knowledge seekers. We believe that every book holds a new world, offering opportunities for learning, discovery, and personal growth. That’s why we are dedicated to bringing you a diverse collection of books, ranging from classic literature and specialized publications to self-development guides and children's books. More than just a book-buying platform, we strive to be a bridge connecting you with timeless cultural and intellectual values. With an elegant, user-friendly interface and a smart search system, you can quickly find the books that best suit your interests. Additionally, our special promotions and home delivery services help you save time and fully enjoy the joy of reading. Join us on a journey of knowledge exploration, passion nurturing, and personal growth every day! ebookbell.com

Compositional Semantics An Introduction To The Syntaxsemantics Interface Pauline Jacobson

  • 1.
    Compositional Semantics AnIntroduction To The Syntaxsemantics Interface Pauline Jacobson download https://ebookbell.com/product/compositional-semantics-an- introduction-to-the-syntaxsemantics-interface-pauline- jacobson-4768428 Explore and download more ebooks at ebookbell.com
  • 2.
    Here are somerecommended products that we believe you will be interested in. You can click the link to download. Introduction To Semantics An Essential Guide To The Composition Of Meaning Thomas Ede Zimmermann Wolfgang Sternefeld https://ebookbell.com/product/introduction-to-semantics-an-essential- guide-to-the-composition-of-meaning-thomas-ede-zimmermann-wolfgang- sternefeld-50987382 Introduction To Semantics An Essential Guide To The Composition Of Meaning Thomas Ede Zimmermann https://ebookbell.com/product/introduction-to-semantics-an-essential- guide-to-the-composition-of-meaning-thomas-ede-zimmermann-4323978 An Introduction To Lexical Semantics A Formal Approach To Word Meaning And Its Composition First Published Eunhee Lee https://ebookbell.com/product/an-introduction-to-lexical-semantics-a- formal-approach-to-word-meaning-and-its-composition-first-published- eunhee-lee-48249722 Presupposition And Implicature In Compositional Semantics Uli Sauerland https://ebookbell.com/product/presupposition-and-implicature-in- compositional-semantics-uli-sauerland-5360714
  • 3.
    Default Semantics FoundationsOf A Compositional Theory Of Acts Of Communication Illustrated Edition K M Jaszczolt https://ebookbell.com/product/default-semantics-foundations-of-a- compositional-theory-of-acts-of-communication-illustrated-edition-k-m- jaszczolt-1394278 The German Perfect Its Semantic Composition And Its Interactions With Temporal Adverbials 1st Edition Renate Musan Auth https://ebookbell.com/product/the-german-perfect-its-semantic- composition-and-its-interactions-with-temporal-adverbials-1st-edition- renate-musan-auth-4481838 Service Composition For The Semantic Web Foreword By Schaharam Dustdar 2011th Edition Brahim Medjahed https://ebookbell.com/product/service-composition-for-the-semantic- web-foreword-by-schaharam-dustdar-2011th-edition-brahim- medjahed-55302634 Agentbased Semantic Web Service Composition Springerbriefs In Electrical And Computer Engineering 2012th Edition Kumar https://ebookbell.com/product/agentbased-semantic-web-service- composition-springerbriefs-in-electrical-and-computer- engineering-2012th-edition-kumar-55302636 Semantic Web Services And Web Process Composition First International Workshop Swswpc 2004 San Diego Ca Usa July 6 2004 Revised Selected Papers 1st Edition Jorge Cardoso https://ebookbell.com/product/semantic-web-services-and-web-process- composition-first-international-workshop-swswpc-2004-san-diego-ca-usa- july-6-2004-revised-selected-papers-1st-edition-jorge-cardoso-1549518
  • 6.
    OXFORD TEXTBOOKS INLINGUISTICS Compositional Semantics
  • 7.
    OXFORD TEXTBOOKS INLINGUISTICS PUBLISHED Compositional Semantics An Introduction to the Syntax/Semantics Interface by Pauline Jacobson The Grammar of Words An Introduction to Linguistic Morphology Third edition by Geert Booij A Practical Introduction to Phonetics Second edition by J. C. Catford Meaning in Language An Introduction to Semantics and Pragmatics Third edition by Alan Cruse Natural Language Syntax by Peter W. Culicover Principles and Parameters An Introduction to Syntactic Theory by Peter W. Culicover A Semantic Approach to English Grammar by R. M. W. Dixon Semantic Analysis A Practical Introduction by Cliff Goddard Pragmatics by Yan Huang The History of Languages An Introduction by Tore Janson Diachronic Syntax by Ian Roberts Cognitive Grammar An Introduction by John R. Taylor Linguistic Categorization Third edition by John R. Taylor IN PREPARATION The Lexicon An Introduction by Elisabetta Ježek Functional Discourse Analysis by Evelien Keizer Semantics and Pragmatics Meaning in Language and Discourse Second edition by Kasia M. Jaszczolt Translation Theory and Practice by Kirsten Malmkjaer Linguistic Typology Theory, Method, Data by Jae Jung Song
  • 8.
    Compositional Semantics An Introductionto the Syntax/Semantics Interface Pauline Jacobson 1
  • 9.
    3 Great Clarendon Street,Oxford, OX2 6DP, United Kingdom Oxford University Press is a department of the University of Oxford. It furthers the University’s objective of excellence in research, scholarship, and education by publishing worldwide. Oxford is a registered trade mark of Oxford University Press in the UK and in certain other countries © Pauline Jacobson 2014 The moral rights of the author have been asserted First Edition published in 2014 Impression: 1 All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, without the prior permission in writing of Oxford University Press, or as expressly permitted by law, by licence, or under terms agreed with the appropriate reprographics rights organization. Enquiries concerning reproduction outside the scope of the above should be sent to the Rights Department, Oxford University Press, at the address above You must not circulate this work in any other form and you must impose this same condition on any acquirer Published in the United States of America by Oxford University Press 198 Madison Avenue, New York, NY 10016, United States of America British Library Cataloguing in Publication Data Data available Library of Congress Control Number: 2013957565 ISBN 978–0–19–967714–6 (Hbk.) 978–0–19–967715–3 (Pbk.) Printed and bound by CPI Group (UK) Ltd, Croydon CR0 4YY Links to third party websites are provided by Oxford in good faith and for information only. Oxford disclaims any responsibility for the materials contained in any third party website referenced in this work.
  • 10.
    To the memoryof my parents, Florence and Nathan Jacobson (Florie and Jake)
  • 12.
    Contents Sections and exercisesmarked with * are more advanced and/or less central (and on some occasions intended as open-ended speculation). These may be skipped with little or no consequence for later sections. Acknowledgments xv Foreword: On Using this Book xvii Part I: Foundational Concepts: Building a Fragment 1. Introduction 3 1.1. Goals 3 1.1.1. Compositional semantics and (some of) the goals of semantic theory 4 1.1.2. Direct Compositionality—and its role in this text 8 1.2. A brief note on the history of semantics within modern linguistic theory 10 1.3. The notion of a “fragment” and its use in this text 12 *1.4. An intriguing puzzle 13 1.5. Appendix: Sets and Functions 19 1.5.1. Sets, members, and subsets 19 1.5.2. Union, intersection, and complement 21 1.5.3. Ordered pairs, relations, equivalence relations, and partitions 21 1.5.4. Functions 24 2. Semantic foundations 27 2.1. Model-theoretic semantics 27 2.2. Truth conditions 28 2.3. Possible worlds 31 2.3.1. Introducing the notion 31 2.3.2. Characteristic function of a set 35
  • 13.
    2.3.3. Notation andterminology 36 2.3.4. Talking about worlds 37 2.4. Times 41 3. Compositionality, Direct Compositionality, and the syntax/semantics interface 42 3.1. Building a fragment: First steps 43 3.2. Implicatures vs truth conditions 46 3.3. Folding in worlds (and times) 48 3.4. Negation: A first pass 49 4. Expanding the fragment: Syntactic categories and semantic types 52 4.1. Noun phrases 52 4.2. Intransitive verbs/verb phrases 58 4.3. A brief look at lexical semantics 60 4.4. Back to the compositional semantics 62 4.5. Illustrating the syntactic/semantic composition 63 5. Transitive verbs: Resolving an apparent syntax/semantics mismatch 66 5.1. First pass: Straw Man’s solution 67 5.2. Abandoning Direct Compositionality 70 5.3. Hidden S conjunction? 73 5.4. Back to the drawing board: Curry’ed functions 76 5.4.1. Recasting the meaning of transitive verbs 76 *5.4.2. Currying more generally 79 5.4.3. Solving Straw Man’s problems 80 *5.5. A note on three-place verbs 84 6. Categorial Grammar 88 6.1. Basics 89 6.2. Syntax/semantics correspondence 92 6.3. Refinements 94 *6.4. Some further observations about syntax: Case marking and agreement 96 6.5. Further generalizing the rules 98 6.6. and, or, negation, and case-marking prepositions 100 6.7. Summary: The fragment so far 104 viii CONTENTS
  • 14.
    7. The autonomyof syntax? 107 7.1. Good syntax—bad semantics 108 7.2. Bad syntax—good semantics 111 8. Adjectives, nouns, determiners, and more 113 8.1. Predicative adjectives 113 8.2. Nouns (“common nouns”) 116 8.3. the 118 8.4. Prenominal adjectives and unary rules 120 8.4.1. An additional binary rule? 120 8.4.2. Interlude: Unary rules (“type shift” rules; morphological rules) 121 8.4.3. Back to adjectives 124 8.4.4. Are all noun modifier adjectives intersective? 126 8.5. Relational nouns 128 8.6. Summary: The fragment so far 131 9. Interlude: The semantics of variables and the lambda calculus 133 9.1. First-order predicate logic (modified) 134 9.1.1. Variables 134 9.1.2. Quantifiers 138 9.2. The lambda calculus 144 9.2.1. Lambda abstraction 144 9.2.2. Lambda conversion 147 9.3. Naming model-theoretic objects 151 Part II: Enriching the Domain 10. Returning to English: Generalized quantifiers 157 10.1. “NPs” with quantificational determiners 157 10.1.1. First pass: Individuals? 158 10.1.2. Second pass: Sets of individuals? 159 10.1.3. Third pass: English as first-order logic? Straw Man makes a cameo reappearance 160 10.2. Generalized quantifiers 162 10.2.1. The semantics 162 10.2.2. The syntax 165 10.2.3. Extending to the other quantified cases 166 CONTENTS ix
  • 15.
    10.3. The determiners167 10.4. Interlude: Scalar implicature 168 10.5. Generalized quantifiers and conjunction 173 10.5.1. VP conjunction 173 10.5.2. Generalized quantifier conjunction 175 *10.6. Domain restriction 178 11. Ordinary NPs and type lifting 181 11.1. Montague’s solution 182 11.2. Partee and Rooth’s solution: Lift as a unary rule 185 11.3. Generalizing the rule 186 11.4. The lowest types hypothesis 189 11.4.1. Some additional predictions 189 *11.4.2. The lowest types hypothesis: Processing vs grammar 192 11.5. Resolving some syntactic loose ends 194 11.6. Summary: The fragment so far 196 12. Generalized conjunction 198 12.1. Generalizing and and or 198 12.2. Remarks on noun conjunction 201 12.3. Summary: The fragment so far 203 Part III: Relative Clauses, Scopes, and Binding: Some Theoretical Controversies 13. Relative clauses: Sketching two accounts 207 13.1. Preliminaries 207 13.2. What do relative clauses modify? 209 13.2.1. The noun modification argument 209 *13.2.2. But—not quite so fast . . . 213 13.3. An account with movement and traces: Assignment functions in grammar 214 13.4. A Direct Compositional (and variable-free) analysis 221 13.4.1. Preliminaries: Function composition 221 13.4.2. Interlude: Coordination 224 13.4.3. Returning to relative clauses 227 13.4.3.1. The basic account 227 13.4.3.2. Syntactic loose ends 231 x CONTENTS
  • 16.
    13.4.4. A refinement:The Geach rule 233 13.4.4.1. The rule and its application to relative clauses (and coordination) 233 *13.4.4.2. Generalized and as generalized g-sl? 238 13.5. Appendix: Reassuring the Appalled Syntactician 240 13.6. Summary: The fragment so far 242 14. Generalized quantifiers in object position: Two approaches 244 14.1. A tempting but inadequate solution 245 14.2. Quantifier (Lowering)/Raising 248 14.3. A Direct Compositional solution 255 14.4. Some comparisons 263 14.4.1. The generality of the solutions 263 14.4.2. Antecedent-Contained Ellipsis 265 14.4.2.1. The “textbook” argument 265 14.4.2.2. Antecedent-Contained Ellipsis as Transitive Verb Phrase Ellipsis 270 14.5. Generalized quantifiers in embedded sentences 272 14.6. Summary: The fragment so far 273 15. The interpretation of pronouns: Two accounts 274 15.1. Introductory remarks 274 15.1.1. Variables or not? 274 15.1.2. Pronominal binding: The empirical phenomenon 276 15.2. The LF (variable-ful) view 278 15.2.1. The basic account 278 15.2.2. A wrinkle: Weak Crossover effects 282 15.2.3. Free pronouns 285 15.3. A Direct Compositional and variable-free approach 286 15.3.1. Pronouns and “binding” 287 15.3.2. Free pronouns 293 *15.3.3. Generalizing the rules 294 15.3.4. Weak Crossover 298 15.3.5. What happens when there is more than one pronoun? 301 CONTENTS xi
  • 17.
    15.3.5.1. Multiple pronounswhich are “the same” 301 *15.3.5.2. And yet another possibility 303 15.3.5.3. More than one pronoun— different binders 304 15.4. More on Weak Crossover 305 15.4.1. Definite NPs and WCO 305 15.4.1.1. An apparent problem (for both theories) 305 15.4.1.2. Not a problem after all 306 15.4.1.3. Independent evidence: Ellipsis and the notion “sloppy identity” 307 *15.4.1.4. A more formal account of the strict reading 312 *15.4.2. A note on three-place verbs and WCO interactions 315 15.5. More on free “variables” 317 15.5.1. Free “variables” in the variable-ful account 317 15.5.2. Free “variables” in the variable-free theory 319 15.6. The i-within-i effect revisited: A solution 320 Appendices to Parts I–III: The full fragment 323 Appendix 1: The basic fragment 323 Foundational concepts 323 The lexicon 324 Word order rules 326 Unary and combinatory rules 327 *Appendix 2: Possible extensions of the fragment for three-place verbs using Wrap 329 The lexicon 329 Word order rule 330 Revision of the definition of functions corresponding to categories of the form A/B 330 Appendix 3: Informally discussed addition for VP Ellipsis 330 xii CONTENTS
  • 18.
    Part IV: FurtherTopics 16. Negative Polarity Items, semantic strength, and scalar implicature revisited 333 16.1. The notion of Negative Polarity Items 334 16.1.1. any and ever 334 16.1.2. Minimizers 336 16.2. The distribution of NPIs 337 16.2.1. First pass 337 16.2.2. Second pass 339 16.3. The first breakthrough 340 16.3.1. Initial characterization of the environments 340 16.3.2. A note on many and few 347 16.4. But why? And how? 349 16.4.1. The insight from semantic strength 351 16.4.2. The second breakthrough: Kadmon and Landman on any 353 16.4.3. ever 356 16.5. Open questions and further discussion 358 16.6. Semantic strength and scalar implicatures 360 17. More binding phenomena 364 17.1. Functional questions 364 17.1.1. The phenomenon 364 17.1.2. An analysis with complex traces 366 17.1.3. A variable-free analysis 368 17.2. Binding and coordination (in variable-free semantics) 370 17.3. “Paycheck” pronouns (and “donkey” pronouns?) 371 17.3.1. Paychecks 371 17.3.2. A variable-free analysis of paycheck pronouns 374 17.3.3. Donkeys 376 17.4. “Hidden variables” 378 17.4.1. Hidden variables can always be complex 378 17.4.2. Translating into variable-free 382 18. Additional semantic dimensions: The semantics of focus 385 18.1. Computing alternatives compositionally 386 18.2. The meaning of only 390 CONTENTS xiii
  • 19.
    18.3. Additional focus-sensitivewords 392 18.4. Further compositional issues 393 18.5. Focus and binding 395 18.6. More generally: Multidimensional semantics 397 19. Intensionality and the syntax/semantics interface 398 19.1. Intensionality-sensitive words 399 19.2. Revising the fragment: Type-sensitive rules 403 References 407 Index 419 xiv CONTENTS
  • 20.
    Acknowledgments This book beganlife as a series of handouts for my Formal Semantics course in 1998. The handouts grew and were revised over the years until I realized they were trying to become a book. I owe my greatest debt to the students in these classes. I never would have undertaken to write this book were it not for the enthusiasm and insight of my students who have always made teaching Formal Semantics just plain fun. I am also indebted to three scholars whose work has influenced my thinking in ways sometimes obvious and sometimes subtle. The influence of Barbara Partee will be obvious throughout this book. To David Dowty, I owe an understanding of how to think “semantically,” and my thinking about Categorial Grammar and its elegance as a theory of the syntax/semantics interface owes much to the work of Emmon Bach. The way of thinking that I learned directly or indirectly from these three semanticists has shaped in various ways the point of view taken in this book, and the way the material is put together. Oxford University Press provided me with three detailed and extraordin- arily helpful referee reports. Probably none of the referees will be satisfied with all of my decisions, but hopefully each will find ways in which their thoughtful comments have greatly improved the manuscript. I thank also Chris Kennedy both for very helpful comments on the manuscript and for “dry-running” parts of it in his courses. I owe a huge debt to Peter Klecha, who did a heroic job of going through the manuscript in detail—working through all of the exercises to make sure they were doable (indeed sometimes they weren’t), to make sure they were not presupposing material not yet introduced (indeed, sometimes they were), and to make sure the formalism was consistent (indeed, sometimes it wasn’t). I have been amazed at Peet’s ability to see just what a student new to the material could or could not be expected to do, and at his astute reading of the material which saved me from some embarrassing mistakes. I am also extremely indebted to Jackson Golden, who has done a fantastic
  • 21.
    job in helpingto put together the bibliography, and to Christopher Mescher for help in compiling the index. I thank NSF for funding on my grant BCS 0646081; although this is a textbook it also contains much of my own research which was funded under the grant. Finally, I thank Michael Rosen, for always encouraging this project, for putting up with its writing over the last two years, and for everything else. xvi ACKNOWLEDGMENTS
  • 22.
    Foreword: On Usingthis Book Each instructor will no doubt find her or his own most useful path through this book (or parts of it), but some suggestions might be helpful. The book obviously contains more material than can realistically be covered in a single semester’s introductory formal semantics course. In teaching this material myself in introductory formal semantics courses (with a mix of graduate students and advanced undergraduates) I have found that I can teach most of the material through Part III plus Chapter 16 and a brief foray into one other of additional topics in Part IV, although I do not go through the material in the full detail given here. (One reason for writing such a book is so that students can get more detail on their own.) Hence, an instructor might choose to do (most of) Chapters 1–15, and one or two additional chapters in Part IV. One plan for this is to spend about three to four weeks on the material through Chapter 5, one week on Chapters 6 and 7 com- bined, and then approximately one chapter a week for the rest (Chapter 12 could take less than a week and Chapter 15 probably more). Depending on the students’ background, another reasonable semester’s goal might instead be to work in detail through Part II and then approach some of the material in Part III. The material would also probably work as the full text for a two- quarter or one-year semantics course (perhaps supplemented at the end with some readings from original sources), and the later material can be the basis of a more advanced course, supplemented with readings from the literature. I hope that the book might also be used by students already familiar with basic formal semantics but not familiar with the viewpoint of Direct Com- positionality (or not familiar with Categorial Grammar); such a student can systematically work through the material from what would be for them a new point of view. This could easily be part of a second-semester or second- year semantics course. There are some choice points in the order in which to read the material. I myself always teach Chapter 16 (on Negative Polarity Items) right after Chapter 10. This is because once students have learned about generalized quantifiers, they have the tools to approach the domain of Negative Polarity Items. I feel it is an especially satisfying reward—after working through
  • 23.
    much technical apparatus—toapply the apparatus to a domain that is rich and so beautifully illustrates the relevance of the formal tools to the distri- bution of items that are entirely natural and conversational. So right after Chapter 10 there is a fork point: one can either go on to Chapter 11 or digress and read Chapter 16. I have set it up so that neither chapter (nor subsequent ones) presupposes the other except in occasional minor ways. Chapter 16 could also be read between Parts II and III. Moreover, the chapters in Part IV can be read in any order. The decision to include two theoretical points of view—the Direct Com- positional view and a view that instead uses Logical Form—poses some difficulties. While I would love to have written the book purely from the point of view of Direct Compositionality, I feel it would be doing students a disservice to not be able to read and appreciate the rich body of work that uses the notion of a Logical Form. My solution, then, was to illustrate these two different viewpoints side by side with respect to a key body of data (relative clauses, quantifiers in object positions, scopes, and binding). But this could be confusing for a student new to this material, so I have set it up in such a way that a student can—on a first pass—work through the relevant chapters carefully from just one of the points of view, and later go back to work through the other. In general, the material on the other point of view will need to be skimmed as it will sometimes contain empirical data of relevance to the discussion, but the technical details can be postponed. Another difficult decision centered on the material on binding. As is pointed out in the text, I give two versions of pronominal binding: a fairly standard one that makes use of Logical Form and variables, and one making no use of variables in the semantics set within the Direct Compos- itional framework. (This is based in part although not exclusively on my own work on variable-free semantics.) But this makes it appear as if Direct Compositionality is crucially tied in with a variable-free semantics. It is not, and in an earlier draft (as well as in the way I have often taught this material) I also exposit an intermediate position: one which is Direct Compositional but does make use of variables. In the end, I decided that to develop three approaches to a single domain was bound to be just too much for an introductory book. With some regret, then, I decided it was prudent to remove the discussion of the intermediate position. Still, I feel it is worth- while to be able to access this material—both for the sake of historical accuracy (it was a well-developed position) and to understand that the issue of whether or not to have variables is to some extent independent of xviii FOREWORD: ON USING THIS BOOK
  • 24.
    the issue ofDirect Compositionality. For those who are not fans of the variable-free approach, I don’t want to leave the impression that if variable- free semantics is wrong, so is Direct Compositionality: there is a very good Direct Compositional semantics of pronouns using variables. So in order to allow access to a discussion of that intermediate position, I have made it available on the textbook website, which is <http://sites.clps.brown.edu/ compositional.sem>. One further decision that some have found unusual is to delay the formal introduction of the lambda calculus until Chapter 9 which is the end of Part I. There is a reason for this. I want to stress that this is just a convenient notation for writing model-theoretic objects, and not some level of repre- sentation that is a crucial part of the theory. (For example, lambda conver- sion is not intended as a rule of grammar, but just as a way to convert one way of writing a model-theoretic object into another way of writing it.) To stress the difference between the notation and the actual meanings of expressions, I have always found it helpful to try to use ordinary prose in naming model-theoretic objects wherever possible, even when this prose leads to cumbersome descriptions like “the set of all sets that contain the dog-set as a subset.” Eventually, of course, English prose gets too cumber- some (and is not sufficiently unambiguous) and we need a better notation (hence the lambda calculus). But I find that insisting on prose whenever possible alongside with a clearer notation is helpful—not only for under- standing that notation is just that, but also to get a better intuition about fancy objects like sets of sets or functions from sets of sets to truth values. However, an instructor who prefers to make use of the lambda calculus earlier can move Chapter 9; it could have been placed after Chapter 6. For the student with no prior background in elementary set theory, I have included an appendix to Chapter 1 that contains all of the basic notions of set theory, ordered pairs, relations, and functions that are necessary for the later material. There is quite a bit of material there, and so I would recom- mend that a student with no relevant background just read the discussion of sets first, and then come back to subsequent notions as they appear in the text. It is often easiest to absorb this material when one has a reason to do so, and when it is in the context of the fuller semantic theory for which it is being pressed into service. A word about the sections and exercises marked with an asterisk. These are more advanced and any of them truly can be skipped (especially on a first pass through the material). They are intended to provide greater coverage and depth for the ambitious and curious student, and the starred FOREWORD: ON USING THIS BOOK xix
  • 25.
    exercises can bequite challenging. Some—especially toward the end—are actually open-ended; I hope that a few might even spawn some interesting research topics for students. In any case, the starred exercises are there to be enjoyed and not to be frustrating, so I would recommend picking and choosing wisely according to a student’s level of interest and ambition. I do hope that some of these will be fun. Finally, as noted above, there is a website for this book at <http://sites.clps. brown.edu/compositional.sem>. It contains the supplementary material on binding and will be updated from time to time with other supplementary material, along with errata as they are discovered. xx FOREWORD: ON USING THIS BOOK
  • 26.
  • 28.
    1 Introduction 1.1. Goals 3 1.2.A brief note on the history of semantics within modern linguistic theory 10 1.3. The notion of a “fragment” and its use in this text 12 *1.4. An intriguing puzzle 13 1.5. Appendix: Sets and Functions 19 1.1. Goals This book stems from a belief that linguistic semantics is a beautiful field, that the tools used to study formal semantics have yielded a rich body of results about fascinating and subtle data, that the field continues to produce exciting new insights at an impressive rate,1 and that there are simple and 1 Readers wishing a taste of many of the ongoing developments in formal semantics and in the syntax/semantics interface might want to look at the journals Linguistics and Philosophy (Springer), Natural Language Semantics (Springer), Journal of Semantics (Oxford University Press), and Semantics and Pragmatics (online journal, available at <http://semprag.org/>), among many other journals. Regular conferences at which cutting-edge research is presented include the annual Semantics and Linguistic Theory (SALT) conference, Sinn und Bedeutung (also annual), the biannual Amsterdam Colloquium for Language, Logic, and Informa- tion, and Semantics of Underrepresented Languages of the Americas, as well as most of the more general regular linguistics conferences. Of course, most of the work in these venues will not be accessible to a student just learning formal semantics, but it is hoped that this book will give a large part of the necessary background for following at least some of this research. In any case, a glance at the list of papers in
  • 29.
    elegant tools tomodel how the syntax and semantics of a natural language work together. We begin with a very elementary “fragment” of English and proceed to expand it further and further—adding tools as needed but aiming to keep the basic machinery relatively simple. The goal of proceeding in this way is to account for a domain of data which is sufficiently rich as to show the excitement of studying formal semantics and its interaction with syntax. We note one limitation from the outset: this book concentrates entirely on the analysis of English. The project of modeling the semantics and the syntax/semantics interaction of any single language already provides such a rich set of results that one can hopefully find this limitation justified for an introductory book like this. In fact, the results that have been gleaned from a detailed modeling of one language have in recent years allowed the field to expand so as to provide a wealth of analyses of other languages.2 This book hopes to give the foundation to approach that literature. 1.1.1. Compositional semantics and (some of) the goals of semantic theory One of the most striking and fundamental properties of language—any language—is that speakers have the ability to produce and understand an unlimited number of expressions that they have never produced or heard before (indeed many of these will have never before been uttered by any- one). This simple point is stressed in just about every introductory linguistics textbook, often phrased this way: “A speaker of a language is able to recognize as well-formed an unlimited number of expressions that s/he has never heard before.” Examples that demonstrate this are easy to construct. One can, for instance, note the existence of expressions like those in (1): any of these venues can give the reader a taste of the richness of the domain of inquiry within linguistic semantics. 2 Much cross-linguistic semantic work can be found in the journals and confer- ence proceedings cited in footnote 1. An early edited volume on this is Bach et al.’s Quantification in Natural Languages. There is now also an annual conference Semantics of Underrepresented Languages of the Americas with published confer- ence proceedings. And many of the specialized conferences on individual languages and language families regularly include work on semantics. 4 1. INTRODUCTION
  • 30.
    (1) a. thetallest linguistics major b. the tallest linguistics major who is graduating in December c. the tallest linguistics major who is graduating in December who is enrolled in formal semantics d. the tallest linguistics major who is graduating in December who is enrolled in formal semantics who took phonology last semester . . . One can keep forming longer and longer expressions like this by adding new relative clauses (each of the phrases that begin with who here is what is commonly known as a relative clause). But while this is often put in terms of a speaker’s ability to recognize that these are well-formed, that is surely only part of the story. Even more interesting (at least to a semanticist) is the fact that speakers know how to interpret these expressions. The rule system that speakers have unconsciously learned is hardly just a system to determine whether a given string of words is an expression of the language in question (here English); language would be quite useless if it were just a collection of meaningless strings. And so, in modeling what a speaker of English “knows” (in an uncon- scious sense, of course) about her/his language we want to predict how it is that s/he can understand expressions like those in (1) no matter how many relative clauses they contain. Thus speakers obviously have as part of their knowledge a finite set of basic items—call these the words and call the collection of the basic items the lexicon. (Here and for most of this text we ignore the distinction between words and morphemes.) Since the lexicon is finite, the meanings of the basic items can be learned on a case-by-case basis. But this obviously cannot be the case for the larger expressions: there has to be some systematic set of principles that speakers have that allows them to understand their meanings on the basis of the meanings of the smaller parts (ultimately the words) that make them up. This is the system which is called the compositional semantics—and one of the jobs of a theory of the seman- tics (of any language) is to model the rules and/or principles which allow speakers to understand an unlimited number of expressions. This book is primarily about just this. Let’s look a bit more at the expressions in (1). When a speaker utters any of these expressions—perhaps as part of a fuller sentence like in (2)—the act of uttering these expressions takes place in a fuller discourse context, and we understand them relative to facts about that context: (2) We need to make sure to order academic regalia which is long enough to fit the tallest linguistics major (who is graduating in December (who . . . )) 1.1. GOALS 5
  • 31.
    The role ofcontext will be discussed more formally at various points in the text, but the informal notion of a speech or discourse context is clear enough. So suppose we are using the expressions in (1) in a context in which it is obvious that we are concerned with the students at Brown University. Given this (or any other context), we can see that any speaker of English immediately knows some interesting facts about these expressions—facts which our model of the compositional semantics needs to account for. Take for instance (1a). It refers to some unique individual.3 The hearer may well not know who exactly that is—in fact the speaker might not either (as is clear in a context like (2)). But both parties assume that there is a particular individual (and only one) referred to by each of these expressions. And there are many other inferences that can be drawn from these. For example, we immediately know that if the individual described by (1a) is Nora, then either she’s also the individual described by (1b) or else she is not graduating in December. Moreover, if Nora is not the person picked out by (1b) then whoever that person is, s/he must be shorter than Nora. Similarly, with each successively longer phrase we either refer to the same person, or to one who is shorter. Suppose that Zorba is the person described by (1b). We know that he is shorter than Nora, and also know that if he is not the person described by (1c) then he is not enrolled in formal semantics. And whoever the (1c) person is—let’s say Otto—Otto must be shorter than Zorba. The addition of each successive relative clause either keeps the referent constant or allows shorter and shorter people to “rise to the top.” This kind of knowledge is automatic and immediate, and it is the job of a model of the compositional semantics to explicitly account for inferences like this. We won’t give a serious account of any of this at this point, but can hint at one possible account. Suppose that an expression like linguistics major refers to some set of individuals. (Readers not familiar with basic notions of set theory should consult the Appendix to this chapter.) When this set is put together with the tallest (pretend that the tallest is a single word here), the entire expression ends up referring to the tallest member of that set. Nothing 3 In reality there could conceivably be two individuals of exactly the same height. But use of the expressions in (1) does seem to assume that there is a unique referent for these. This is sometimes called a presupposition; these are rather odd expressions if the speaker knows that there are two individuals with exactly the same height (in that case the speaker might have said the two tallest linguistics majors). 6 1. INTRODUCTION
  • 32.
    surprising so far.But what is more interesting is what happens with the addition of further relative clauses. It seems plausible that something like who is graduating in December also refers to a set (obviously, the set of December graduates). The above facts will make sense if the compositional semantics first combines the two sets (the set of linguistics majors and the set of December graduates) and intersects them to give a new set. (The intersection of two sets is all things that are in both sets; again see the Appendix.) So (1b) ends up picking out the tallest member of that set. It is now possible to demonstrate that the system correctly predicts that if the referent of (1b) is not Nora, it can only be because she is not graduating in December. For if Nora is taller than anyone in the linguistics major set (call that L) then she is taller than anyone in the intersection of L with the December graduates (call that D). After all, everyone who is in that intersection of L and D is also in L. So if Nora is not the referent of (1b) it can only be that she’s not in the intersection of D and L, and since she’s in L (by assumption) it follows that she can’t be in D. It also follows that if (1b) refers to Zorba, he must be shorter than Nora. By the definition of intersection, if Zorba is in the intersection of D and L he is in L, but we already know that Nora is taller than everyone else in L. All of this is very simple logic that we—the linguists—can work out in the form of an informal proof as above. It could also be worked more formally if one were so inclined. Pedantic though it may seem, it shows that our compositional procedure (which involves intersecting two sets) can be used to correctly model inferences that speakers of English effort- lessly make. Moreover, the appeal here is that this is perfectly general and extends no matter how many new relative clauses are added. Take (1c). The semantics set-up above extends immediately to this. The new relative clause in (1c) is who is enrolled in formal semantics. This picks out yet another set—and so this now intersects with the set that we already formed for (1b). The fact that the referent of (1c) can either be Zorba or someone shorter than Zorba follows by the same logic shown above; the reader can work out the details. And the procedure can be repeated over and over no matter how many relative clauses are introduced. 1.1. GOALS 7
  • 33.
    1.1.2. Direct Compositionality—andits role in this text This book has a rather ambitious set of goals. On the one hand, I intend this to be a stand-alone text for anyone wishing to have an introduction to formal semantics, compositional semantics, or what is commonly known as the syntax/semantics interface. In other words, we will be asking (as in the above example) what a compositional semantics might look like: how can we model the tools available (again, of course, unconsciously) to speakers of a language that allow them to compute meanings of larger expressions from the meanings of the smaller ones that make them up. What are the formal ways in which meanings combine? And what are the types of objects that we need in order to model that? (For example, the discussion above shows that some simple tools of set theory can be useful.) But while most semanticists agree that (in general) the meaning of a larger expression is built in some systematic way from the meanings of the parts that make it up, just exactly how the syntactic system of a language and the compositional semantics work together is a matter of considerable controversy, and is one of the central questions addressed in this book. And so this book takes one particular point of view on this: the point of view known as Direct Compositionality. This view was explored perhaps most notably in Montague (1970) and was either generally accepted or at least taken as a serious desideratum in much of the work in linguistic formal semantics throughout the 1970s and 1980s (particularly work in what was then known as the Montague Grammar program). It was also taken as the foundation for semantics in syntactic theories such as Generalized Phrase Structure Grammar (Gazdar, Klein, Pullum, and Sag 1985), and is assumed in a large body within current grammatical theories that go under the rubric of Categorial Grammar, Type-Logical Grammar, and other related theories. 1.1. Because this example is just meant to illustrate the notion of a compositional semantics, we have made some assumptions about the order in which the semantics put things together without justifying them. Suppose that rather than the way it was set up here, the meanings of the two relative clauses (1c) first combined, and then that combined with linguistics major. Would that make any difference to the basic semantic compositional picture that we have set up here? Would the procedure extend correctly to (1d)? 8 1. INTRODUCTION
  • 34.
    To elucidate, afairly uncontroversial claim is that the grammar of any natural language is a system of rules (or principles, if one prefers) that define the set of well-formed expressions of the language (i.e., the syntax) and a set of rules (or principles) pairing these with meanings (i.e., the semantics). The hypothesis of Direct Compositionality is a simple one: the two systems work in tandem. Each expression that is proven well-formed in the syntax is assigned a meaning by the semantics, and the syntactic rules or principles which prove an expression as well-formed are paired with the semantics which assign the expression a meaning. (An interesting consequence of this view is that every well-formed syntactic expression does have a meaning.4 ) It is not only the case that every well-formed sentence has a meaning, but also each local expression (“constituent”) within the sentence that the syntax defines as well-formed has a meaning. Of course putting it this way is arguably not much more than just a slogan: the empirical content of this depends in part on just how the syntax works and what one takes to be a meaning. This will be filled in as we proceed. It might also seem at first glance that the hypothesis of Direct Compositionality is a fairly trivial one. But in fact it is not always immediately obvious how to give a Direct Compositional analysis. Even the example in 1.1.1 is a case in point. If the syntax and semantics work together, then the analysis given above leads to the conclusion that in the syntax a relative clause like who is graduating in December combines with linguistics major rather than with the tallest lin- guistics major. But this very question regarding the syntax of relative clauses has been debated in the literature since the 1960s, and many researchers have claimed that the syntactic constituent structure of the tallest linguistics major who is graduating next year is not the structure that was used above for the semantic analysis. We will actually revisit this particular question in later chapters (see, e.g., section 13.2). So one of the goals of this book will be to see what it takes to give Direct Compositional analyses of a variety of constructions. While the material in this book is generally exposited from the Direct Compositional point of view (along with discussion of the challenges to this hypothesis), the book is also intended to be a perfectly reasonable 4 Of course one of the earliest arguments in Generative Grammar for divorcing the syntax from the semantics (and thus a putative argument against Direct Com- positionality) is based on the claim that there are well-formed expressions that don’t have any meaning (Chomsky 1957). This is addressed in section 7.1. 1.1. GOALS 9
  • 35.
    stand-alone textbook forany formal semantics course. Thus it is suitable for any linguistics student or linguist wanting a ground-up introduction to formal semantics, and for a philosophy or logic student wanting a background in formal semantics within linguistics. In the service of being a stand-alone text in modern formal semantic theory, the book will, where relevant, also develop the mechanics of at least one fairly standard non- Direct Compositional theory of the syntax/semantics interface. This is done especially in Part III (Chapters 13–15) where some phenomena are discussed from both direct and non-Direct Compositional points of view. There are several reasons for expositing parallel Direct and non-Direct Compositional accounts of some domains. One is to enable readers to approach the wide range of literature written from either point of view. Second, this allows for a serious comparison of two different approaches. Third, learning more than one set of details for the analysis of any construction allows for a deeper understanding of the basic generalizations and results— generalizations which often transcend the particulars of one theoretical implementation. Finally, a student who has already learned formal seman- tics from a non-Direct Compositional point of view can hopefully also profit from this book by seeing an interesting fragment of English explicitly analyzed from the Direct Compositional point of view. 1.2. A brief note on the history of semantics within modern linguistic theory The subfield of semantics as a core field in modern linguistic theory is relatively recent and is one of the fastest growing subfields.5 Early work within the general enterprise of generative grammar had little to say about semantics. To be sure, by the end of about the 1960s and the early 1970s there was considerable discussion as to how the syntax and the semantics interacted; such discussion was mostly framed in terms of a debate between Generative Semantics (see, e.g., McCawley 1971; Lakoff 1971) and Inter- pretive Semantics (see Chomsky 1970; Jackendoff 1972). We will not discuss the content of that debate here, but much of the work framed within these 5 A much more extensive and authoritative history of the development of formal semantics within modern linguistic theory can be found in Partee (forthcoming). 10 1. INTRODUCTION
  • 36.
    two competing pointsof view did not incorporate a systematic view of the semantics itself. Of course, the linguistics literature during that period contained many seminal observations about semantic notions such as scope, negation, and “binding,” but these were generally not embedded within a full-blown theory of semantics designed to capture semantic notions like entailment and truth conditions (see Chapter 2), although they easily could have been embedded into such a theory. The fact that semantics was not taken a subfield in and of itself during this period comes—at least in part— from Noam Chomsky’s emphasis on syntax during the early development of generative grammar. Chomsky (1957) explicitly rejects the notion that semantics is relevant in the construction of grammars, and this notion persisted for quite some time. It is probably fair to say that modern formal semantics as a subfield within linguistic theory began in the early to mid-1970s with the cross-fertilization of linguistic theory and philosophy of language (including semantics) sparked by Barbara Partee, Richmond Thomason, David Lewis, and others. Par- tee’s work was a particularly influential bridge between linguistics and philosophy as she had originally been a student of Chomsky’s at MIT and always had a strong interest in the connections between language and logic, and hence in topics like quantifiers, negation, etc. As an assistant professor at UCLA, she became acquainted with the seminal work of Richard Mon- tague, a philosopher and logician who (among his many other contributions within philosophy and logic) had a major interest in modeling the semantics (and the syntax/semantics interaction) of natural language (although Montague himself dealt only with English). In fact, the program of Direct Compositionality is advocated in his work (see especially Montague 1970). We will have more to say about his specific contributions as this book proceeds; for now, we note that one of the appeals of his work from the point of view of a linguist was his notion that the semantic composition of natural language reflects and respects its syntax. Partee saw the relevance of Montague’s work to linguistic theory and wrote a series of papers aimed at synthesizing some of the insights from Montague’s work with results within Transformational Grammar (see, for example, Partee 1973). At the same time, the appearance of Lewis (1970), Stalnaker and Thomason (1973), and other work in the philosophy of language also helped launch modern formal semantics and cement its connection to linguistic theory. Such work within philosophy as well as Partee’s early group of students (both at UCLA and later at the University of Massachusetts) continued the tradition, 1.2. THE HISTORY OF SEMANTICS 11
  • 37.
    broadening the domainof inquiry and results. From there was born the enterprise known as Montague grammar6 which eventually gave rise to the more general subfield of formal semantics. Montague himself died in 1971,7 and the field of formal semantics evolved in many ways quite different from the original work in Montague grammar. Nonetheless, many of the basic tools of linguistic formal semantics as it is developed to this day stem from some of this early work cited above. Since the late 1970s the field has blossomed, and is now within linguistics generally considered as one of the core areas along with at least phonology and syntax. 1.3. The notion of a “fragment” and its use in this text Inspired by the work of Montague in papers such as Montague (1973), much work in formal semantics within the 1970s and 1980s took it as axiomatic that a goal was to formulate fully explicit grammars (in both syntactic and semantic detail) of the fragment of the language one is concerned with (English in most such work). The term “fragment” got extended to mean not only the portion of the language being modeled, but also the portion of the grammar being proposed as an explicit account of the facts. The strategy of writing fragments (of grammars) has the advantage of giving an explicit theory which makes testable predictions, and of making theory and/or proposal comparison easier. Unfortunately, the goal of formulating fully explicit fragments went out of style during the last two decades or so. This is in part due to the fact that linguistic theories often promised that many of the particular details did not need to be stated as they would fall out from very general principles. It is certainly reasonable to hope that this is ultimately true, but the relevant principles often go unstated or are stated only rather vaguely, making it extremely difficult to really compare proposals and/or evaluate theories and theoretical claims. Having rules and principles be as general as possible is, of course, highly desirable. But this does not mean that they should not be 6 An excellent introduction to the general program of Montague semantics and an explication especially of Montague (1973) can be found in Dowty, Wall, and Peters (1981). 7 Montague was murdered on March 7, 1971. No arrest was ever made in conjunction with the murder. 12 1. INTRODUCTION
  • 38.
    formulated explicitly—only thatmore mileage will be gotten out of explicit formulations. The present text is therefore committed to trying to revive the notion of explicit fragment construction. We cannot promise to give every detail of the domain of English syntax and semantics we are trying to model. Some parts will be left tentative, some stated informally, and some simply omitted. Nonetheless, the goal is to give a reasonable amount of an explicit fragment. We will therefore periodically take stock by summarizing the fragment constructed so far, and a full summary is provided at the end of Part III. *1.4. An intriguing puzzle This introductory chapter concludes with an illustration of a puzzle, a solution to which is proposed in section 15.6. However, the goal here is not to champion any one particular solution, and readers may safely skip this section and return to the data only in section 15.6. But we include this in the introductory remarks for the reader who wants a preview of just what kinds of complex and subtle data a theory of syntax and semantics ultimately hopes to account for. To fully appreciate the particular puzzle here, one should keep the following in mind. The contrasts are quite real; the judgments have been checked with many speakers over the years by myself and many others. Yet—like other subtle facts in syntax, phon- ology, and semantics—these are not generalizations which we have ever been consciously taught nor even generalizations that most of us are even aware of until we see them in a linguistics course (or book). What, then, is there about our unconscious knowledge of the grammatical system that predicts these judgments? This is the sort of puzzle that theories of semantics and its interaction with syntax ultimately seek to solve. So, consider what we will call the A-B-C party scenario. I go to a small party consisting of only myself and three married couples: Alice and Abe, Betty and Bert, and Cathy and Carl. I learn that Alice and Abe met each other only a few years ago, and similarly for Cathy and Carl. But interest- ingly, I also find out that Betty and Bert have been sweethearts since childhood. I like Betty a lot, and spend a good part of the evening talking to her. The next day, you ask me how I enjoyed the party and if there was anyone that I especially enjoyed meeting. I certainly can answer with (3): *1.4. AN INTRIGUING PUZZLE 13
  • 39.
    (3) Oh yes,I especially enjoyed talking to Betty. But now suppose that I can’t remember Betty’s name, although I do remem- ber that her husband’s name is Bert. I can answer with either (4a) or (4b): (4) a. Oh yes, I especially enjoyed talking to—oh, I can’t remember her name— you know, the woman who is married to Bert. b. Oh yes, I especially enjoyed talking to—oh, I can’t remember her name— you know, the wife of Bert. ((4b) would sound more natural if we substituted Bert’s wife for the wife of Bert; this will not impact on the ultimate point and the exposition is simplified using (4b).) Now, let us tweak the scenario slightly and assume that I am one of those people who just doesn’t remember names very well. As a result, I remember neither Betty’s name nor Bert’s name, although I do remember the interest- ing fact that they are the only couple at the party who have been sweethearts since childhood. As an answer to your question, (5) would be quite natural: (5) Oh yes, I especially enjoyed talking to—oh, I can’t remember her name—you know, the woman who is married to her childhood sweetheart. But what is striking is that I can’t answer with (6): (6) *Oh yes, I especially enjoyed talking to—oh I can’t remember her name—you know, the wife of her childhood sweetheart. (We are taking liberties with the * notation here. This is generally used in works in syntax to indicate something that is ill-formed. (6) in fact is fine, just not on the intended reading; and we will continue to notate a sentence with an asterisk in front of it when we mean “bad on a particular reading” provided that it is clear what the intended reading is.) It should be noted that some speakers find the contrast rather subtle but there is general agreement that (6) is stranger than (5). All of these examples contain various extra material (the parentheticals, etc.) which are there to make them sound natural and conversational. But as we proceed it will be convenient to strip away the parts that are irrelevant to figuring out the semantics, so we can recast (6) as (7)—also impossible as an answer to the question and in this context: (7) *I especially enjoyed talking to the wife of her childhood sweetheart. 14 1. INTRODUCTION
  • 40.
    Of course, (7)is a perfectly good sentence, but it cannot be used in our party scenario as a way to identify Betty. Since some readers do find the contrast subtle, two points are worth noting. First, one should resist the temptation to recast (7) in one’s mind as I especially enjoyed talking to the one who’s the wife of her childhood sweetheart or I especially enjoyed talking to the woman who’s the wife of her childhood sweetheart. That would be cheating; the point is not to find a closely related way to say the same thing but to notice that the actual way in (7) contrasts with I especially enjoyed talking to the woman who is married to her childhood sweetheart (and contrasts with the above variants too). As to why these variants are good, we return to that shortly. Moreover, while the contrasts above may be subtle for some speakers, there is a related mystery where the facts dramatically pop out. Thus take (8) in the same scenario, where the only people at issue are Alice, Betty, and Cathy: (8) Betty is the only woman who is married to her childhood sweetheart. This can be making two different claims. The obvious one in this scenario is that Cathy is not married to Cathy’s childhood sweetheart, and Alice is not married to Alice’s childhood sweetheart. The other is the “non-polygam- ous” reading: it asserts that Bert (or whoever Betty’s husband might be) has only one wife. Since we (generally) assume that people have just one wife, this reading (given standard assumptions) is not the first one that someone would think of since it is less likely to be conveying any interesting information. But despite the fact that the non-polygamous reading is the less obvious one for (8), it is the only reading (or at least the one that pops out first) for (9): (9) Betty is the only wife of her childhood sweetheart. Why should that be? We’ll put (8) and (9) aside for the moment, and return to the simpler case of (7). So the mystery is why (7) is bad as a way to identify Betty. This is especially puzzling in that both (4b) and (5) are perfectly good—or, to give their stripped-down versions, (10) and (11) are both fine. Each one differs minim- ally from our bad case, yet neither of these two has any problem. (10) I especially enjoyed talking to the woman who is married to her childhood sweetheart. (11) I especially enjoyed talking to the wife of Bert. *1.4. AN INTRIGUING PUZZLE 15
  • 41.
    So surely thereis nothing incoherent or wrong with the meaning that (7) is trying convey, for (10) is just a slightly different form and conveys exactly this meaning. Hence the puzzle has something to do with the mapping between syntax and semantics: why one is a good way to package the relevant information while the other is not. We can informally recast the puzzle in the following way. Compare the two expressions the woman who is married to Bert and the wife of Bert. (Following a long tradition within linguistics, we will refer to these as NPs, which comes from “noun phrases.” They are also in much modern literature referred to as DPs, for “determiner phrases,” but we stick to the more traditional terminology in this text.)8 Both of these can correspond to meanings that we can (roughly and informally) represent as (12): (12) the x: x is a woman and x is married to Bert But while the object NP in (10) can be represented as in (11), the object NP in (7) cannot: (13) the x: x is a woman and x is married to x’s childhood sweetheart The basic phenomenon here was discussed in, among others, Jacobson (1977) (where it was called Langendoen’s constraint), Chomsky (1981 under the rubric of i-within-i condition), and many since. As there seems to be nothing wrong with the meaning, we can assume that the phenomenon in question has something to do with the way the syntax and semantics interact. Notice that we have given a kind of formula (and one that uses a “variable” x) to represent the meanings in question, but for now we should think of these simply as placeholders to bring out the intuition. After all, recasting her in the above examples as x doesn’t really immediately give us the tools for computing the meanings of the expression: we have traded a 8 In theories which use the term DP, the NP is used instead to refer to material after the Determiner; e.g., mother of Romeo in an expression like the mother of Romeo. Here we will be calling this simply N (i.e., a “noun”) and allowing terms like N to refer both to simple material consisting of just one word and to complex material. This is discussed further in Chapter 6. We are aware that this will initially cause some confusion to a reader who is used to using “NP” to mean a noun and its complement, but it is well worth becoming fluent in both sets of terminologies. The terminology here is the traditional one found in large amounts of literature not only in syntax and semantics but in neighboring fields like psycholinguistics, philosophy, cognitive science, etc. 16 1. INTRODUCTION
  • 42.
    pronoun her fora variable x. But this accomplishes little until we have a way to think about what a variable like x means. (Indeed this is explored in detail in Chapters 9 and 15, including developing an alternative view that does not make use of variables in the semantics.) We thus caution that formulas like (12) and (13) are best seen simply as informal and helpful ways to bring out the intended meanings. Similarly, one often sees indices used in the literature as a way to bring this out; one will find discussions using the notation in (14) and (15) to make the point, where the indexation in (14) indicates a good possible reading for the NP while (15) cannot be understood in the intended way: (14) the womani whoi is married to heri childhood sweetheart (15) *the wifei of heri childhood sweetheart Much work in grammatical theory actually assumes that NPs and pronouns come with (obviously silent) indices in the syntax; here we will be using indices from time to time simply as a way to notate intended readings without any commitment to their being actual pieces of grammatical machinery. Before leaving this (for now), there’s one other interesting point to notice. However we ultimately state the principle, the claim is that an NP like the wife of her childhood sweetheart cannot correspond to the meaning shown earlier in (13): (13) the x: x is a woman and x is married to x’s childhood sweetheart But one might think that this is not really correct, since it is in fact just fine to use (16) as a way to identify Betty: (16) I especially enjoyed meeting the woman who is the wife of her childhood sweetheart. This point was made earlier; many speakers on reading (7) tend to recast it in their minds as (16). Similarly, (17) is impeccable on the understanding where her is Betty: (17) Betty is the wife of her childhood sweetheart. But a close reflection reveals that this does not threaten the generalization. Again, using indices or variables simply as a convenient way to elucidate the point, it is easy to see in (17) that her need not be “coindexed with” (or “correspond to the same variable as”) wife but rather it just refers directly to Betty. That is, we can represent it as in (18a) using indices, or as in (18b) using the more spelled-out formula (though still quite informal). *1.4. AN INTRIGUING PUZZLE 17
  • 43.
    (18) a. Bettyjis [NP the wifei of herj childhood sweetheart]. b. Betty, y [y = the x: x is a woman and x is married to y’s childhood sweetheart] Since we are asserting identity between Betty and the person married to Betty’s childhood sweetheart, it of course follows that Betty is married to Betty’s childhood sweetheart and so the full sentence (16) will end up with the relevant meaning.9 But the claim that the object NP itself (the wife of her childhood sweetheart) does have the meaning represented in (12) is not threatened. The same point holds for (17), whose meaning can be represented as (19a) or (19b). (19) a. the womanj whoj is the wifei of herj childhood sweetheart b. the y: y is a woman and y = the x: x is a woman and x is married to y’s childhood sweetheart Is there a way to confirm that this is the right sort of explanation for these apparent counterexamples? Indeed there is, and it centers on the contrast between (8) and (9) which was discussed earlier. We leave it to the interested reader in the exercise to play with this and get a sense of why (8) is ambiguous and (9) is not. Having completed that, one should be able to see how it is that this gives support for the explanation offered above as to why (17) does not threaten the claim that the wife of her childhood sweetheart cannot correspond to the meaning shown informally in (12). *1.2. Work out—using the informal representations either with indices or the representations with variables—why it is that (7) is ambiguous and (8) is not. Of course you will need to think a bit about how to treat only, but nothing very complex is required. You can be perfectly informal in your treatment of only, but you should be able to get a feel for why these two differ. 9 This general observation—although for a slightly different case—was made in Postal (1970) who distinguished between “presuppposed” coreference and “asserted” coreference. Here the fact that Betty and the wife of her childhood sweetheart end up “referring” to the same individual is exactly what the sentence is asserting. 18 1. INTRODUCTION
  • 44.
    As noted atthe outset of this section, the goal here is just to provide a mystery to whet the reader’s appetite; the tools needed to provide a hypoth- esis as to the explanation of the mystery will be developed later. 1.5. Appendix: Sets and Functions 1.5.1. Sets, members, and subsets Since the notions of sets and of functions are crucial throughout this book, some formal definitions and discussion are provided here for readers not entirely familiar with these notions. We begin with the notion of a set. A set is simply any collection of objects (it can have a finite number of objects, an infinite number, or none at all). For example, we can talk about the set of positive integers less than 10; sets can be notated by listing the members and enclosing the list in curly brackets: {1,2,3,4,5,6,7,8,9}. The order in which they are listed makes no difference; a set is just a collection of things without any order. So if we were to write {2,5,3,4,9,7,8,1,6}, this names the same set. Each integer in this set is called a member or an element of the set. If we were to name this set A, then the notation 4 ∈ A means that 4 is a member (or element) of A. Something either is or is not in a set; it makes no sense to say it occurs twice (or more) in the set. Note also that a set can have a single member; this is called a singleton set. Thus {4} is the set with only one member; this set is distinct from 4 itself. (4 is a member of {4}.) A set can have an infinite number of members; the set of positive integers for example is infinite. Obviously this can’t be named by listing the mem- bers. One can in this case specify the set by a recursive procedure. Call the set I, then one can specify I by two statements: (a) (what is known as the base step): 1 ∈ I, and (b) (the recursion step) if n ∈ I then n+1 ∈ I. (It is understood when one lists things this way that nothing else is in I.) One will also often see a notation which describes rather than lists the members. For example, we can write the following set, call it B: {x|x is a New England state}. This names a finite set, and so we could also give B in list form as follows: {Maine, New Hampshire, Vermont, Massachusetts, Rhode Island, Con- necticut}. These two are just different notations for naming the same set. This can also be used, of course, for infinite sets. Take, for example, the set 1.5. APPENDIX: SETS AND FUNCTIONS 19
  • 45.
    {x|x is aninteger and x > 9}. This names the set of integers greater than 9. And, a set can have no members. There is only one such set; its name is the null set or the empty set, and is generally written as . Of course, there are other ways one can describe the null set. For example, the set of integers each of which is greater than 9 and less than 10 is the empty set. The cardinality of some set refers to the number of elements in that set; the notation |B| means the cardinality of B. Hence, given our set B above, |B| is six. Take some set A. Then a subset of A is any set all of whose members are also in A. Suppose, for example, we begin with a set C which is {1,2,3}. Then {1,2} is a subset of C, as is {1,3} and so forth. The notation for the subset relation is . The full definition of subset is as follows: B A if and only if every member of B is a member of A. From this it follows that every set is a subset of itself (so for the set C above, one of its subsets is the set {1,2,3}). It is, however, sometimes convenient to refer to those subsets distinct from the original set; in that case we can talk about a proper subset of some set. The symbol for this is , so B ` if and only if B A and B 6¼ A. Since the definition of subset says that B is a subset of A if and only if everything that is in B is also in A, it follows that if nothing is in B then B is a subset of A. thus the null set is a subset of every other set. Sets themselves can have sets as members, and so one can talk about the set of all subsets of a set A. This is called the power set of A, written as P (A). For example, given the set C above, P (A) = {, {1}, {2}, {3}, {1,2}, {1,3}, {2,3}, {1,2,3}}. We will also have occasion to talk about the reverse of the subset relation—i.e., the superset relation. A is a superset of B if and only if B is a subset of A. The notation for this is A B. Once again this is defined in such a way that every set is a superset of itself; a superset of B which is not identical to B is called a proper superset, and the notation for this is . *1.3. If a set A has n members, then the number of subsets of A is 2n . Try to see why this is true. Hint: for every member x of some set A, then for each subset B of A, x is either in B or is not in B. 1.4. How many members does the following set have: {}? *1.5. What is P ()? 20 1. INTRODUCTION
  • 46.
    1.5.2. Union, intersection,and complement Take any two sets A and B. Then there is a set C which consists of everything that is in A and everything that is in B. This is called the union of A and B, and is written A [ B. For example, if A is {1,2,3} and B is {2,4,6} then A [ B is {1,2,3,4,6}. Moreover, for any two sets A and B the intersection of A and B is the set of all things that are in both A and B. This is written A B. So, for example, in the case directly above, the intersection of A and B is {2}. Or, if we were to intersect the set of integers which can be evenly divided by 2 (the set of even integers) with the set of integers which can be evenly divided by 3, we end up with the set of integers that can be evenly divided by 6. One final useful notion here is the complement of a set. The complement of some set A is the set of all things which are not in A (this is sometimes notated as A0 ). Usually one talks about this notion with respect to some larger domain. Strictly speaking, the complement of {1,2,3} would include not only all integers greater than 3 but also all sorts of other numbers (like 1/3), the sun, my dog Kiana, and the kitchen sink. Rarely are we interested in that sort of set; so in practice when one talks about “the complement of some set A” this is generally with respect to some larger set B of which A is a subset. Then the complement of A refers to all things in B that are not in A (this is notated as B-A). For example, when restricting the discussion to the set of positive integers, the complement of {1,2,3} is the set of all integers greater than 3. 1.5.3. Ordered pairs, relations, equivalence relations, and partitions Sets are unordered collections of objects. But it is quite useful (as will become very apparent as this book proceeds) to be able to talk about pairs of objects that are ordered in some way. An ordered pair is just that: 1.6. a. For any two sets A and B such that A B, what set is A [ B? b. For any two sets A and B such that A B, what set is A B? 1.5. APPENDIX: SETS AND FUNCTIONS 21
  • 47.
    it is twoobjects with some ordering between them. If the two objects are a and b, then (a,b) is an ordered pair; (b,a) is a different ordered pair. An ordered pair need not contain distinct items: (a,a) is an ordered pair. In applying this to actual natural relations that exist in the world we are generally interested in sets of ordered pairs. (One can generalize this notion to ordered triples and so forth; an ordered n-tuple means an ordered list of n items.) This notion is easiest to grasp with some concrete examples. Take again the set {1,2,3}, and take the relation “is greater than.” Then this can be seen as a set of ordered pairs; if we are restricting this to items from our little 1-2-3 set, this would be the set {(2,1), (3,1), (3,2)}. Now suppose we instead take the following set of ordered pairs: {(2,1), (3,1), (3,2), (1,1), (2,2), (3,3)}. Then (restricting this again to our 1-2-3 set) we have now actually listed the relation “is greater than or equal to.” Or, take the set {(1,1), (2,2), (3,3)}. That is the relation “is equal to” (defined for the set of integers {1,2,3}). In other words, what we are calling a relation is just some set of ordered pairs. In the example above, both the first and second member of each ordered pair was drawn from the same set (the set {1,2,3}). But this is not necessary; we can have a set of ordered pairs each of whose first member is drawn from some set A and the second member from some set B where A and B are different (they can, but need not, have some of the same members). For example, the relation “is the capital of” is a relation between cities and states; it can be expressed as a set of ordered pairs of the general form {(Providence, Rhode Island), (Boston, Massachusetts), (Springfield, Illinois), (Pierre, South Dakota), . . . } (the . . . here is a shorthand for the remaining 46 pairs). Take two sets A and B (they could be the same set or different). Then A x B refers to the set of all ordered pairs whose first member is in A and whose second member is in B. (This is also called the Cartesian product of A and B.) As in the case above, it is helpful to give the intuition of this by coming up with some concrete example. Suppose we take as our set A some group of professors—say, Professor Magoo, Professor Carberry, and Professor Glazie. Call that set P (for shorthand, let’s call its members m, c, and g, so P is the set {m,c,g}). Now suppose we have a set S which consists of three students who we will just indicate as x, y, and z (so S = {x,y,z}). Then P x S = {(m,x}, (m,y), (m,z), (c,x), (c,y), (c,z), (g,x), (g,y), (g,z)}. Suppose that Magoo wrote a letter of recommendation for all three students, Carberry wrote one for only y, and Glazie wrote one for y and z. Then the relation “wrote a 22 1. INTRODUCTION
  • 48.
    recommendation for” isa subset of P x S, and is the set of ordered pairs {(m,x}, (m,y), (m,z), (c,y), (g,y), (g,z)}. More generally, we define a relation (between members of A and B) as any subset of A x B. There are some special and interesting properties that can be defined when the two sets are the same. That is, we are now looking at subsets of A x A. Consider a relation R (some subset of A x A) which is such that for all x in A, (x,x) is in R. Such a relation is called a reflexive relation. (These need not be the only kinds of pairs to be in R for R to be reflexive; other pairs can be in there too.) For example, if talking about the set of integers again, the relation “is greater than or equal to” is reflexive; for all numbers n, (n,n) is in the set of ordered pairs described by that relation. A relation R is called irreflexive if for all x in A, (x,x) is not in R. Further, consider any two members x and y (both members of A). Then if it’s the case that for all x and y if (x,y) is in R then (y,x) is also in R, the relation is called symmetric. Imagine, for example, a lovely world with no unrequited love. Then is in love with is symmetric in that world. If our set were {m, c, g, and p}, then if the pair (m,c) were in our relation R (i.e., “is in love with”) the fact that R is symmetric means that (c,m) is also in R. (Notice that our definition neither requires (c,c) to be in R nor excludes that; either is possible.) Or, to look at a relation which is symmetric by definition: consider the relation is a sibling of. (While is a sibling of is symmetric, is a sister of is not. Why not?) One final useful definition is a transitive relation. A transitive relation R is one for which for every x, y, and z, if (x,y) is in R and (y,z) is in R, then (x,z) is in R. (The relation “is greater than” is transitive, as is the relation “is greater than or equal to”). Any relation R which is reflexive, transitive, and symmetric is called an equivalence relation. As an example of such a relation, consider the set of students (call it S) at an elementary school that services grades 1 through 6. Then “is in the same grade as” is an equivalence relation in S x S. (While it is unusual to use the phrase “in the same grade as” when referring to the same person it seems false to say Johnny is not in the same grade as himself so we can see that this relation is reflexive.) It is also obvious that it is symmetric and transitive. Note that this—and any other equivalence relation—divides up the original set (here, S) into a group of non-overlapping subsets. The set of these subsets is called a partition. Thus, a partition of any set S is a set of subsets of S such that for each distinct subset A and B, A B = , and the union of all the subsets is S. To show that any equivalence relation induces such a partition, take any x in S and define Sx as {y|(y,x) is in R}. Since R is 1.5. APPENDIX: SETS AND FUNCTIONS 23
  • 49.
    reflexive, we knowthat x is in Sx (and hence we know that Sx is guaranteed not to be empty). Moreover, the fact that R is reflexive means that each member of S is guaranteed to be in at least one such subset, so we know that the union of all of these is S. We can further show that for any two such subsets Sa and Sb, they either have no members in common (i.e., they have a null intersection) or they are the same. Thus, take any c which is in both Sa and Sb. By definition, this means that (c,a) is in R and (c,b) is in R. By the fact that R is transitive and symmetric, it follows that (a,b) and (b,a) are in R (the reader can work through the necessary steps). But then, for all x such that (x,a) is in R, (x,b) is also in R. To show this note again that R is transitive. If (x,a) is in R and (a,b) is in R then (x,b) is also in R. Hence given the initial premise that there is a non-empty intersection between Sa and Sb, it follows that everything in Sa is in Sb. That everything in Sb is also in Sa follows in the same way, and so the two are the same set. Each subset in a partition is called a cell in that partition. In the example above, the cells correspond to the different grades. (There don’t have to be six cells—it could be that one of the grades has no student in it. But there can be no more than six; recall that by definition a cell can’t be empty.) Just as any equivalence relation induces a partition, given any partition one can give an equivalence relation that corresponds to any partition; this is the relation that holds between any two a and b in S such that a and b are in the same cell in the partition. 1.5.4. Functions A function takes every member of some set A and assigns it a value from a set B (B could be the same set as A, but need not be). This can also be formalized using the notion of a set of ordered pairs. Thus, consider two sets A and B (which again could be the same but need not be). Then, a (total) function from A to B is any set of ordered pairs (i.e., any subset of A x B) such that for each a in A, there is one and only one ordered pair with a as first member. Thus if we think of the function f as assigning to each a in A some values in B, note that the criterion above ensures that each member of A is indeed assigned a value, and is assigned a unique value. A is referred to as the domain of the function, and B is referred to as the co-domain. For any function f and any a in the domain of f, we write f(a) to indicate the value that f assigns to a. (To use other common terminology, f(a) means 24 1. INTRODUCTION
  • 50.
    the result thatone gets by applying the function f to a.) There is no restriction that each member of B must appear as second member of some ordered pair; the term range of the function f is the set of all b in B such that there is some a such that f(a) = b. Note that these definitions are such that the range of a function is a subset of the co-domain. In practice (at least in works within linguistics) the terms “range” and “co-domain” are often not distinguished. As noted above, there is no restriction that each member of B appear as second member of an ordered pair. Nor is there a restriction that it appear only once. If each member of B is used as a value only once (that is, for each b in B, there is a unique a such that f(a) = b) then B obviously can be no smaller than a. It can have more members, or it can be the same size. If the latter is the case, then it also follows that for every b in B, there is some a such that f(a) = b. When both conditions above hold (i.e., for each b in B, there is one and only one a such that f(a) = b, we say that there is a one-to- one correspondence between A and B. Note that for any function f which is a one-to-one correspondence, there is a corresponding function f-1 which is just the reverse: it is a function mapping each member of B to a member of A such that for all a in A and b in B, if f(a) = b then f-1 (b) = a.10 We will have some occasion to talk about the notion of a partial function. A partial function is one where not every member of A is actually assigned a value by f; f is undefined for some subset of A. (Of course any partial function f is also a total function with a smaller domain.) We can illustrate this by returning to our earlier example of ordered pairs of US cities and states, where the first member of each ordered pair is the capital of the second. This is a partial function from the set of US cities to states (not every US city is a capital). We can reverse it, and have each state as the first member of the ordered pair and the second as its capital (this function could be expressed in prose as has as its capital). This is now a total function from 10 Incidentally, the notion of the availability of a one-to-one correspondence can be used to define what it means for two sets to have the same cardinality. Obviously for two finite sets it is clear what it means to have the same cardinality, since we can count the members. But consider the case of infinite sets. Take the following two sets: A = the set of positive integers {1,2,3, . . . } and B = the set of positive even integers {2,4,6, . . . }. Both are infinite. Surprisingly (when one first hears this) they are also of the same cardinality, because one can establish a one-to-one correspond- ence between them (each member of A is paired with a member of B by multiplying by 2: we will never run out of members in B). 1.5. APPENDIX: SETS AND FUNCTIONS 25
  • 51.
    the set ofstates (every state does have a capital) to the set of US cities. But it is not a one-to-one correspondence for the same reason that our original relation is not a total function; there are many cities without the honor of being a capital. Occasionally in this text it will be useful to list out some actual functions—that is, to name every member in the domain and name what the function at issue maps that member to. There are a variety of ways one could do this. To illustrate, take a domain of four children {Zacky, Yonnie, Shelley, and Baba} (call that set C) and four men {Abe, Bert, Carl, David} (call that set M). Suppose there is a function f from C to M which maps each child to their father. Assume that Abe is the father of Zacky and Yonnie, Bert is the father of Shelley, and David is the father of Baba. Then one can write this information out in various ways. One would be to simply give the set of ordered pairs: {(Zacky, Abe), (Yonnie, Abe), (Shelley, Bert), (Baba, David)}. Usually this notation, however, is not terribly easy to read. We could also write this out in either of the ways shown in (20): (20) a. f(Zacky) = Abe b. Zacky ! Abe f(Yonnie) = Abe Yonnie ! Abe f(Shelley) = Bert Shelley ! Bert f(Baba) = David Baba ! David Or, sometimes it is more convenient to list out the domain on the left and the co-domain on the right and connect them with arrows as in (21): (21) Zacky Abe Yonnie Bert Shelley Carl Baba David Which notation is chosen makes no difference; the choice should be dictated by clarity. 26 1. INTRODUCTION
  • 52.
    2 Semantic foundations 2.1. Model-theoreticsemantics 27 2.2. Truth conditions 28 2.3. Possible worlds 31 2.4. Times 41 2.1. Model-theoretic semantics The primary focus of this book is the syntax/semantics interface—that is, how the syntax and semantics work so that sentences (and other well- formed linguistic expressions) are paired with a meaning. But of course this task is impossible without some idea of what meaning is. Can we talk about meaning without relegating it to the realm of the mysterious, or leaving it solely to folks who work on cognition to deal with? The answer (of course) is yes—there is a rich tradition within linguistics and the philosophy of language for modeling linguistic meaning. In some early work within linguistic theory—especially in the 1960s— meaning was taken to be just a symbolic representation (call it a Logical Form, or LF). While it remains an open question as to whether such representations play an actual role in the way the grammar pairs expressions with meanings, this book (along with much other modern work in formal semantics) assumes that meaning is not just some string of symbols, but rather some actual object out there in the world. Call this a model-theoretic object. (More precisely, we are taking meaning to be an object which forms part of a model which is an abstract representation of the world: hence the term model theory.) Of course, we need some way to name these objects, and so throughout we will use strings of symbols as ways to name them. But the
  • 53.
    point is thatthe grammar maps each linguistic expression into something beyond just a symbolic representation. Otherwise—as so aptly pointed out by David Lewis (1970)—we are simply mapping one language (say, English) into another (what Lewis termed “Markerese”). Yet language is used to convey facts about the world; we draw inferences about the world from what we hear and we gain information about what is true and what is not. So semantics must be a system mapping a linguistic expression to something in the world. But what exactly is meant by model-theoretic objects? These can in fact be quite abstract. Still, they are the “stuff” that is out there in the universe— something constructed out of actual bits of the universe (or, at least, the ontology of the universe as given by language). This would include things like individuals, times, possibilities, and perhaps others; just what are the basic objects that we need is an open question and is part of what semantic theory addresses. The strategy here will be to use a fairly sparse set of primitive objects, and construct more complex objects out of these. Let us, then, begin by setting up two basic building blocks which are foundational in much of the work in linguistic formal semantics. 2.2. Truth conditions A common adage in semantics is: “To know the meaning of a (declarative) sentence is to know what it would take to make it true.” We can use this adage as a first step in constructing the building blocks for meanings: a fundamental fact about declarative sentences is that they are either true or false1 (and since we use language to communicate information about the world, a listener will in general assume that a sentence they have just heard 1 Henceforth we use the term “sentence” to mean a declarative sentence. There is actually no reason to consider questions to be of the same category as declarative sentences even though they also are traditionally referred to as “sentences.” Ques- tions have a different external distribution from declarative sentences (for example, wonder can occur only with a question as its complement, not an ordinary sentence, while the reverse is true for believe); they have a different kind of meaning, and they have a different internal structure. Whether imperatives and declaratives should be considered the same category is a bit less clear (they are more similar), but we will not deal with those here either. 28 2. SEMANTIC FOUNDATIONS
  • 54.
    is true, anduses that fact to enrich their knowledge of the world). Thus (1) is true and (2) is false: (1) Barack Obama moved into the White House on Jan. 20, 2009. (2) John McCain moved into the White House on Jan. 20, 2009. Hence, one basic notion used for the construction of meanings is a truth value—for now assume that there are just two such values: true and false. (More on this directly.) The claim that truth values are a fundamental part of meaning is also motivated by noting that—as shown by the examples above—speakers have intuitions about truth, given certain facts about the world, just like they do about acceptability. And these judgments can be used to test the adequacy of particular theories of meaning. Following standard practice, we use 1 for true and 0 for false. Thus the set of truth values {1,0} and we will also refer to this set as t. Let us use [[Æ]] to mean the semantic value (i.e., the meaning) of a linguistic expression Æ. Then (temporarily) we can say that [[Barack Obama moved into the White House on Jan. 20, 2009]] = 1. Some worries should immediately spring to mind. The most obvious is that something seems amiss in calling the meaning of (1) “true” even if we are willing to accept the fact that it is true. We will enrich the toolbox directly to take care of that. But there are other objections: does it really make sense to say that all declarative sentences are true or false? Clearly not—for some sentences the truth value depends on who is speaking (and on when the sentence is spoken). Take (3): (3) I am President of the United States. This is true if spoken by Barack Obama in 2011, but not if spoken by John McCain and not true if spoken by Barack Obama in 2006. So this has no truth value in and of itself. Nonetheless once certain parameters are fixed (time of utterance and speaker) it is either true or false. So we might want to think of the meaning of (3) as a function into {1,0}—it does yield a truth value but only once we fix certain parameters. But it seems inescapable that a declarative sentence is telling us something about the world, and so truth values are certainly one fundamental piece. In fact, there are many parameters that need to be set in order to derive a truth value. Certain words like I, you, here, now, etc. quite obvi- ously have the property that their value depends on when, where, and by whom these are spoken (these are called indexicals). There are also more 2.2. TRUTH CONDITIONS 29
  • 55.
    Another Random Documenton Scribd Without Any Related Topics
  • 56.
    of such ananimal shows an abundant vegetation, and a climate so mild, that the rivers were not covered with heavy ice in winter; for the supposition that this old hippopotamus was a migratory animal seems very unlikely. Another animal of this time, was the magnificent deer, known as the Irish elk; and which perhaps had its principal abode on the great plain which is now the Irish Sea. The terrible machairodus, or cymetar-toothed tiger, was continued from the Pliocene; and in addition to species of bear still living, there was a species of gigantic size, probably now extinct, the cave bear. Evidences are accumulating, to show that all or nearly all these survived until the human period. If we turn now to those animals which are only locally extinct, we meet with some strange, and at first sight puzzling anomalies. Some of these are creatures now limited to climates much colder than that of Britain. Others now belong to warmer climates. Conspicuous among the former are the musk-sheep, the elk, the reindeer, the glutton, and the lemming. Among the latter, we see the panther, the lion, and the Cape hyena. That animals now so widely separated as the musk-sheep of Arctic America and the hyena of South Africa, could ever have inhabited the same forests, seems a dream of the wildest fancy. Yet it is not difficult to find a probable solution of the mystery. In North America, at the present day, the puma, or American lion, comes up to the same latitudes with the caribou, or reindeer, and moose; and in Asia, the tiger extends its migrations into the abodes of boreal animals in the plains of Siberia. Even in Europe, within the historic period, the reindeer inhabited the forests of Germany; and the lion extended its range nearly as far northward. The explanation lies in the co-existence of a densely wooded country with a temperate climate; the forests affording to southern animals shelter from the cold or winter; and equally to the northern animals protection from the heat of summer. Hence our wonder at this association of animals of diverse habitudes as to climate, is merely a prejudice arising from the present exceptional condition of Europe. Still it is possible that changes unfavourable to some of these animals, were in progress before the arrival of man, with his clearings and forest fires and other disturbing agencies. Even in America, the megalonyx, or gigantic sloth, the mammoth, the mastodon, the fossil horse, and many other creatures,
  • 57.
    disappeared before theModern period; and on both continents the great Post-glacial subsidence or deluge may have swept away some of the species. Such a supposition seems necessary to account for the phenomena of the gravel and cave deposits of England, and Cope has recently suggested it in explanation of similar storehouses of fossil animals in America.[AS] [AS] Proceedings of the American Philosophical Society, April 1871. Among the many pictures which this fertile subject calls up, perhaps none is more curious than that presented by the Post-glacial cavern deposits. We may close our survey of this period with the exploration of one of these strange repositories; and may select Kent’s Hole at Torquay, so carefully excavated and illumined with the magnesium light of scientific inquiry by Mr. Pengelly and a committee of the British Association. The somewhat extensive and ramifying cavern of Kent’s Hole is an irregular excavation, evidently due partly to fissures in limestone rock, and partly to the erosive action of water enlarging such fissures into chambers and galleries. At what time it was originally cut we do not know, but it must have existed as a cavern at the close of the Pliocene or beginning of the Post-pliocene period, since which time it has been receiving a series of deposits which have quite filled up some of its smaller branches. First and lowest, according to Mr. Pengelly, is a “breccia” or mass of broken and rounded stones, with hardened red clay filling the interstices. Most of the stones are of the rock which forms the roof and walls of the cave, but many, especially the rounded ones, are from more distant parts of the surrounding country. In this mass, the depth of which is unknown, are numerous bones, all of one kind of animal, the cave bear, a creature which seems to have lived in Western Europe from the close of the Pliocene down to the modern period. It must have been one of the earliest and most permanent tenants of Kent’s Hole at a time when its lower chambers were still filled with water. Next above the breccia is a floor of “stalagmite” or stony carbonate of
  • 58.
    lime, deposited fromthe drippings of the roof, and in some places three feet thick. This also contains bones of the cave bear, deposited when there was less access of water to the cavern. Mr. Pengelly infers the existence of man at this time from a single flint flake and a single flint chip found in these beds; but mere flakes and chips of flint are too often natural to warrant such a conclusion. After the old stalagmite floor above mentioned was formed, the cave again received deposits of muddy water and stones; but now a change occurs in the remains embedded. This stony clay, or “cave earth” has yielded an immense quantity of teeth and bones, including those of the elephant, rhinoceros, horse, hyena, cave bear, reindeer, and Irish elk. With these were found weapons of chipped flint, and harpoons, needles, and bodkins of bone, precisely similar to those of the North American Indians and other rude races. The “cave earth” is four feet or more in thickness, It is not stratified, and contains many fallen fragments of rock, rounded stones, and broken pieces of stalagmite. It also has patches of the excrement of hyenas, which the explorers suppose to indicate the temporary residence of these animals; and in one spot, near the top, is a limited layer of burnt wood, with remains which indicate the cooking and eating of repasts of animal food by man. It is clear that when this bed was formed the cavern was liable to be inundated with muddy water, carrying stones and other heavy objects, and breaking up in places the old stalagmite floor. One of the most puzzling features, especially to those who take an exclusively uniformitarian view, is, that the entrance of water-borne mud and stones implies a level of the bottom of the water in the neighbouring valleys of about 100 feet above its present height. The cave earth is covered by a second crust of stalagmite, less dense and thick than that below, and containing only a few bones, which are of the same general character with those below, but include a fragment of a human jaw with teeth. Evidently, when this stalagmite was formed, the influx of water-borne materials had ceased, or nearly so; but whether the animals previously occupying the country still continued in it, or only accidental bones, etc., were introduced into the cave or lifted from the bed below, does not appear.
  • 59.
    The next bedmarks a new change. It is a layer of black mould from three to ten inches thick. Its microscopic structure does not seem to have been examined; but it is probably a forest soil, introduced by growth, by water, by wind, and by ingress of animals, at a time when the cave was nearly in its present state, and the surrounding country densely wooded. This bed contains bones of animals, all of them modern, and works of art ranging from the old British times before the Roman invasion up to the porter-bottles and dropped halfpence of modern visitors. Lastly, in and upon the black mould are many fallen blocks from the roof of the cave. There can be no doubt that this cave and the neighbouring one of Brixham have done very much to impress the minds of British geologists with ideas of the great antiquity of man, and they have, more than any other Post-glacial monuments, shown the persistence of some animals now extinct up to the human age. Of precise data for determining time, they have, however, given nothing. The only measures which seed to have been applied, namely, the rate of growth of stalagmite and the rate of erosion of the neighbouring valleys, are, from the very sequence of the deposits, obviously worthless; and the only apparently available constant measure, namely, the fall of blocks from the roof, seems not yet to have been applied. We are therefore quite uncertain as to the number of centuries involved in the filling of this cave, and must remain so until a surer system of calculation is adopted. We may, however, attempt to sketch the series of events which it indicates. The animals found in Kent’s Hole are all “Post-glacial.” They therefore inhabited the country after it rose from the great Glacial submergence. Perhaps the first colonists of the coasts of Devonshire in this period were the cave bears, migrating on floating ice, and subsisting, like the Arctic bear, and the black bears of Anticosti, on fish, and on the garbage cast up by the sea. They found Kent’s Hole a sea- side cavern, with perhaps some of its galleries still full of water, and filling with, breccia, with which the bones of dead bears became mixed. As the land rose, these creatures for the most part betook themselves to lower levels, and in process of time the cavern stood
  • 60.
    upon a hill-side,perhaps several hundreds of feet above the sea; and the mountain torrents, their beds not yet emptied of glacial detritus, washed into it stones and mud and carcases of animals of many species which had now swarmed across the plains elevated out of the sea, and multiplied in the land. This was the time of the cave earth; and before its deposit was completed, though how long before, a confused and often-disturbed bed of this kind cannot tell, man himself seems to have been added to the inhabitants of the British land. In pursuit of game he sometimes ascended the valleys beyond the cavern, or even penetrated into its outer chambers; or perhaps there were even in those days rude and savage hill-men, inhabiting the forests and warring with the more cultivated denizens of plains below, which are now deep under the waters. Their weapons, lost in hunting, or buried in the flesh of wounded animals which crept to the streams to assuage their thirst, are those found in the cave earth. The absence of human bones may merely show that the mighty hunters of those days were too hardy, athletic, and intelligent, often to perish from accidental causes, and that they did not use this cavern for a place of burial. But the land again subsided. The valley of that now nameless river, of which the Rhine the Thames, and the Severn may have alike been tributaries, disappeared under the sea; and some tribe, driven from the lower lands, took refuge in this cave, now again near the encroaching waves, and left there the remains of their last repasts ere they were driven farther inland or engulfed in the waters. For a time the cavern may have been wholly submerged, and the charcoal of the extinguished fires became covered with its thin coating of clay. But ere long it re-emerged to form part of an island, long barren and desolate; and the valleys having been cut deeper by the receding waters, it no longer received muddy deposits, and the crust formed by drippings from its roof contained only bones and pebbles washed by rains or occasional land floods from its own clay deposits. Finally, the modern forests overspread the land, and were tenanted by the modern animals. Man returned to use the cavern again as a place of refuge or habitation, and to leave there the relics contained in the black earth. This seems at present the only intelligible history of this curious cave and others resembling it; though, when we consider the imperfection of the results obtained even by a large amount of labour, and the
  • 61.
    difficult and confusedcharacter of the deposits in this and similar caves, too much value should not be attached to such histories, which may at any time be contradicted or modified by new facts or different explanations of those already known. The time involved depends very much, as already stated, on the question whether we regard the Post- glacial subsidence and re-elevation as somewhat sudden, or as occupying long ages at the slow rate at which some parts of our continents are now rising or sinking.[AT] [AT] Another element in this is also the question raised by Dawkins, Geikie, and others as to subdivisions of the Post-glacial period and intermissions of the Glacial cold. After careful consideration of these views, however, I cannot consider them as of much importance. Such are the glimpses, obscure though stimulating to the imagination, which geology can give of the circumstances attending the appearance of man in Western Europe. How far we are from being able to account for his origin, or to give its circumstances and relative dates for the whole world, the reader will readily understand. Still it is something to know that there is an intelligible meeting-place of the later geological ages and the age of man, and that it is one inviting to many and hopeful researches. It is curious also to find that the few monuments disinterred by geology, the antediluvian record of Holy Scripture, and the golden age of heathen tradition, seem alike to point to similar physical conditions, and to that simple state of the arts of life in which “gold and wampum and flint stones”[AU] constituted the chief material treasures of the earliest tribes of men. They also point to the immeasurable elevation, then as now, of man over his brute rivals for the dominion of the earth. To the naturalist this subject opens up most inviting yet most difficult paths of research, to be entered on with caution and reverence, rather than in the bold and dashing spirit of many modern attempts. The Christian, on his part, may feel satisfied that the scattered monumental relics of the caves and gravels will tell no story very different from that which he has long believed on other evidence, nor anything inconsistent with those views of man’s heavenly origin and destiny which have been the most precious inheritance of the greatest and best minds of every age, from that early pre-historic
  • 62.
    period when men,“palaeolithic” men, no doubt, began to “invoke the name of Jehovah,” the coming Saviour, down to those times when life and immortality are brought to light, for all who will see, by the Saviour already come. [AU] So I read the “gold, bedolah, and shoham” of the description of Eden in Genesis ii.—the oldest literary record of the stone age. In completing this series of pictures, I wish emphatically to insist on the imperfection of the sketches which I have been able to present, and which are less, in comparison with the grand march of the creative work, even as now imperfectly known to science, than the roughest pencilling of a child when compared with a finished picture. If they have any popular value, it will be in presenting such a broad general view of a great subject as may induce further study to fill up the details. If they have any scientific value, it will be in removing the minds of British students for a little from the too exclusive study of their own limited marginal area, which has been to them too much the “celestial empire” around which all other countries must be arranged, and in divesting the subject of the special colouring given to it by certain prominent cliques and parties. Geology as a science is at present in a peculiar and somewhat exceptional state. Under the influence of a few men of commanding genius belonging to the generation now passing away, it has made so gigantic conquests that its armies have broken up into bands of specialists, little better than scientific banditti, liable to be beaten in detail, and prone to commit outrages on common sense and good taste, which bring their otherwise good cause into disrepute. The leaders of these bands are, many of them, good soldiers, but few of them fitted to be general officers, and none of them able to reunite our scattered detachments. We need larger minds, of broader culture and wider sympathies, to organise and rule the lands which we have subdued, and to lead on to further conquests. In the present state of natural science in Britain, this evil is perhaps to be remedied only by providing a wider and deeper culture for our young men. Few of our present workers have enjoyed that thorough
  • 63.
    training in mentalas well as physical science, which is necessary to enable men even of great powers to take large and lofty views of the scheme of nature. Hence we often find men who are fair workers in limited departments, reasoning most illogically, taking narrow and local views, elevating the exception into the rule, led away by baseless metaphysical subtleties, quarrelling with men who look at their specialties from a different point of view, and even striving and plotting for the advancement of their own hobbies. Such defects certainly mar much of the scientific work now being done. In the more advanced walks of scientific research, they are to some extent neutralised by that free discussion which true science always fosters; though even here they sometimes vexatiously arrest the progress of truth, or open floodgates of error which it may require much labour to close. But in public lectures and popular publications they run riot, and are stimulated by the mistaken opposition of narrow-minded good men, by the love of the new and sensational, and by the rivalry of men struggling for place and position. To launch a clever and startling fallacy which will float for a week and stir up a hard fight, seems almost as great a triumph as the discovery of an important fact or law; and the honest student is distracted with the multitude of doctrines, and hustled aside by the crowd of ambitious groundlings. The only remedy in the case is a higher and more general scientific education; and yet I do not wonder that many good men object to this, simply because of the difficulty of finding honest and competent teachers, themselves well grounded in their subjects, and free from that too common insanity of specialists and half-educated men, which impels them to run amuck at everything that does not depend on their own methods of research. This is a difficulty which can be met in our time only by the general good sense and right feeling of the community taking a firm hold of the matter, and insisting on the organization and extension of the higher scientific education, as well as that of a more elementary character, under the management of able and sane men. Yet even if not so counteracted, present follies will pass away, and a new and better state of natural science will arise in the future, by its own internal development. Science cannot long successfully isolate itself from God. Its life lies in the fact that it is the
  • 64.
    exponent of theplans and works of the great Creative Will. It must, in spite of itself, serve His purposes, by dispelling blighting ignorance and superstition, by lighting the way to successive triumphs of human skill over the powers of nature, and by guarding men from the evils that flow from infringement of natural laws. And it cannot fail, as it approaches nearer to the boundaries of that which may be known by finite minds, to be humbled by the contemplation of the infinite, and to recognise therein that intelligence of which the human mind is but the image and shadow. It may be that theologians also are needed who shall be fit to take the place of Moses to our generation, in teaching it again the very elements of natural theology; but let them not look upon science as a cold and godless demon, holding forth to the world a poisoned cup cunningly compounded of truth and falsehood; but rather as the natural ally and associate of the gospel of salvation. The matter is so put in one of those visions which close the canon of revelation, when the prophet sees a mighty angel having the “everlasting gospel to preach;” but he begins his proclamation by calling on men to “worship Him that made heaven and earth and the sea and the fountains of waters.” Men must know God as the Creator even before they seek Him as a benefactor and redeemer. Thus religion must go hand in hand with all true and honest science. In this way only may we look forward to a time when a more exact and large-minded science shall be in perfect accord with a more pure and spiritual Christianity, when the natural and the spiritual shall be seen to be the necessary complements of each other, and when we shall hear no more of reconciliations between science and theology, because there will be no quarrels to reconcile. Already, even in the present chaos of scientific and religious opinion, indications can be seen by the observant, that the Divine Spirit of order is breathing on the mass, and will evolve from it new and beautiful worlds of mental and spiritual existence. CHAPTER XIV.
  • 65.
    PRIMITIVE MAN. CONSIDERED WITHREFERENCE TO MODERN THEORIES AS TO HIS ORIGIN. The geological record, as we have been reading it, introduces us to primitive man, but gives us no distinct information as to his origin. Tradition and revelation have, it is true, their solutions of the mystery, but there are, and always have been, many who will not take these on trust, but must grope for themselves with the taper of science or philosophy into the dark caverns whence issue the springs of humanity. In former times it was philosophic speculation alone which lent its dim and uncertain light to these bold inquirers; but in our day the new and startling discoveries in physics, chemistry, and biology have flashed up with an unexpected brilliancy, and have at least served to dazzle the eyes and encourage the hopes of the curious, and to lead to explorations more bold and systematic than any previously undertaken. Thus has been born amongst us, or rather renewed, for it is a very old thing, that evolutionist philosophy, which has been well characterised as the “baldest of all the philosophies which have sprung up in our world,” and which solves the question of human origin by the assumption that human nature exists potentially in mere inorganic matter, and that a chain of spontaneous derivation connects incandescent molecules or star-dust with the world, and with man himself. This evolutionist doctrine is itself one of the strangest phenomena of humanity. It existed, and most naturally, in the oldest philosophy and poetry, in connection with the crudest and most uncritical, attempts of the human mind to grasp the system of nature; but that in our day a system destitute of any shadow of proof, and supported merely by vague analogies and figures of speech, and by the arbitrary and artificial coherence of its own parts, should be accepted as a philosophy, and should find able adherents to string upon its thread of hypotheses our vast and weighty stores of knowledge, is surpassingly strange. It seems to indicate that the accumulated facts of our age
  • 66.
    have gone altogetherbeyond its capacity for generalisation; and but for the vigour which one sees everywhere, it might be taken as an indication that the human mind has fallen into a state of senility, and in its dotage mistakes for science the imaginations which were the dreams of its youth. In many respects these speculations are important and worthy of the attention of thinking men. They seek to revolutionise the religious beliefs of the world, and if accepted would destroy most of the existing theology and philosophy. They indicate tendencies among scientific thinkers, which, though probably temporary, must, before they disappear, descend to lower strata, and reproduce themselves in grosser forms, and with most serious effects on the whole structure of society. With one class of minds they constitute a sort of religion, which so far satisfies the craving for truths higher than those which relate to immediate wants and pleasures. With another and perhaps larger class, they are accepted as affording a welcome deliverance from all scruples of conscience and fears of a hereafter. In the domain of science evolutionism has like tendencies. It reduces the position of man, who becomes a descendant of inferior animals, and a mere term in a series whose end is unknown. It removes from the study of nature the ideas of final cause and purpose; and the evolutionist, instead of regarding the world as a work of consummate plan, skill, and adjustment, approaches nature as he would a chaos of fallen rocks, which may present forms of castles and grotesque profiles of men and animals, but they are all fortuitous and without significance. It obliterates the fine perception of differences from the mind of the naturalist, and resolves all the complicated relations of living things into some simple idea of descent with modification. It thus destroys the possibility of a philosophical classification, reducing all things to a mere series, and leads to a rapid decay in systematic zoology and botany, which is already very manifest among the disciples of Spencer and Darwin in England. The effect of this will be, if it proceeds further, in a great degree to destroy the educational value and popular interest attaching to these sciences, and to throw them down at the feet of a system of debased metaphysics. As redeeming features in all this, are the careful study of varietal forms, and the inquiries as to the limits of
  • 67.
    species, which havesprung from these discussions, and the harvest of which will be reaped by the true naturalists of the future. Thus these theories as to the origin of men and animals and plants are full of present significance, and may be studied with profit by all; and in no part of their applications more usefully than in that which relates to man. Let us then inquire,—1. What is implied in the idea of evolution as applied to man? 2. What is implied in the idea of creation? 3. How these several views accord with what we actually know as the result of scientific investigation? The first and second of these questions may well occupy the whole of this chapter, and we shall be able merely to glance at their leading aspects. In doing so, it may be well first to place before us in general terms the several alternatives which evolutionists offer, as to the mode in which the honour of an origin from apes or ape-like animals can be granted to us, along with the opposite view as to the independent origin of man which have been maintained either on scientific or scriptural grounds. All the evolutionist theories of the origin of man depend primarily on the possibility of his having been produced from some of the animals more closely allied to him, by the causes now in operation which lead to varietal forms, or by similar causes which have been in operation; and some attach more and others less weight to certain of these causes, or gratuitously suppose others not actually known. Of such causes of change some are internal and others external to the organism. With respect to the former, one school assumes an innate tendency in every species to change in the course of time.[AV] Another believes in exceptional births, either in the course of ordinary generation or by the mode of parthenogenesis.[AW] Another refers to the known facts of reproductive accelleration or retardation observed in some humble creatures.[AX] New forms arising in any of these ways or fortuitously, may, it is supposed, be perpetuated and increased and further improved by favouring external circumstances and the effort of the organism to avail itself of these,[AY] or by the struggle for existence and the survival of the fittest.[AZ] [AV] Parsons, Owen.
  • 68.
    [AW] Mivart, Ferris. [AX]Hyatt and Cope. [AY] Lamarck, etc. [AZ] Darwin, etc. On the other hand, those who believe in the independent origin of man admit the above causes as adequate only to produce mere varieties, liable to return into the original stock. They may either hold that man has appeared as a product of special and miraculous creation, or that he has been created mediately by the operation of forces also concerned in the production of other animals, but the precise nature of which is still unknown to us; or lastly, they may hold what seems to be the view favoured by the book of Genesis, that his bodily form is a product of mediate creation and his spiritual nature a direct emanation from his Creator. The discussion of all these rival theories would occupy volumes, and to follow them into details would require investigations which have already bewildered many minds of some scientific culture. Further, it is the belief of the writer that this plunging into multitudes of details has been fruitful of error, and that it will be a better course to endeavour to reach the root of the matter by looking at the foundations of the general doctrine of evolution itself, and then contrasting it with its rival. Taking, then, this broad view of the subject, two great leading alternatives are presented to us. Either man is an independent product of the will of a Higher Intelligence, acting directly or through the laws and materials of his own institution and production, or he has been produced by an unconscious evolution from lower things. It is true that many evolutionists, either unwilling to offend, or not perceiving the logical consequences of their own hypothesis, endeavour to steer a middle course, and to maintain that the Creator has proceeded by way of evolution. But the bare, hard logic of Spencer, the greatest English authority on evolution, leaves no place for this compromise, and shows that the theory, carried out to its legitimate consequences, excludes the knowledge of a Creator and the possibility of His work. We have, therefore, to choose between evolution and creation; bearing in mind,
  • 69.
    however, that theremay be a place in nature for evolution, properly limited, as well as for other things, and that the idea of creation by no means excludes law and second causes. Limiting ourselves in the first place to theories of evolution, and to these as explaining the origin of species of living beings, and especially of man, we naturally first inquire as to the basis on which they are founded. Now no one pretends that they rest on facts actually observed, for no one has ever observed the production of even one species. Nor do they even rest, like the deductions of theoretical geology, on the extension into past time of causes of change now seen to be in action. Their probability depends entirely on their capacity to account hypothetically for certain relations of living creatures to each other, and to the world without; and the strongest point of the arguments of their advocates is the accumulation of cases of such relations supposed to be accounted for. Such being the kind of argument with which we have to deal, we may first inquire what we are required to believe as conditions of the action of evolution, and secondly, to what extent it actually does explain the phenomena. In the first place, as evolutionists, we are required to assume certain forces, or materials, or both, with which evolution shall begin. Darwin, in his Origin of Species, went so far as to assume the existence of a few of the simpler types of animals; but this view, of course, was only a temporary resting-place for his theory. Others assume a primitive protoplasm, or physical basis of life, and arbitrarily assigning to this substance properties now divided between organised and unorganised, and between dead and living matter, find no difficulty in deducing all plants and animals from it. Still, even this cannot have been the ultimate material. It must have been evolved from something. We are thus brought back to certain molecules of star-dust, or certain conflicting forces, which must have had self-existence, and must have potentially included all subsequent creatures. Otherwise, if with Spencer we hold that God is “unknowable” and creation “unthinkable,” we are left suspended on nothing over a bottomless void, and must adopt as the initial proposition of our philosophy, that all things were made out of nothing, and by nothing; unless we prefer
  • 70.
    to doubt whetheranything exists, and to push the doctrine of relativity to the unscientific extreme of believing that we can study the relations of things non-existent or unknown. So we must allow the evolutionist some small capital to start with; observing, however, that self-existent matter in a state of endless evolution is something of which we cannot possibly have any definite conception. Being granted thus much, the evolutionist next proceeds to demand that we shall also believe in the indefinite variability of material things, and shall set aside all idea that there is any difference in kind between the different substances which we know. They must all be mutually convertible, or at least derivable from some primitive material. It is true that this is contrary to experience. The chemist holds that matter is of different kinds, that one element cannot be converted into another; and he would probably smile if told that, even in the lapse of enormous periods of time, limestone could be evolved out of silica. He may think that this is very different from the idea that a snail can be evolved from an oyster, or a bird from a reptile. But the zoologist will inform him that species of animals are only variable within certain limits, and are not transmutable, in so far as experience and experiment are concerned. They have their allotropic forms, but cannot be changed into one another. But if we grant this second demand, the evolutionist has a third in store for us. We must also admit that by some inevitable necessity the changes of things must in the main take place in one direction, from the more simple to the more complex, from the lower to the higher. At first sight this seems not only to follow from the previous assumptions, but to accord with observation. Do not all living things rise from a simpler to a more complex state? has not the history of the earth displayed a gradually increasing elevation and complexity? But, on the other hand, the complex organism becoming mature, resolves itself again into the simple germ, and finally is dissolved into its constituent elements. The complex returns into the simple, and what we see is not an evolution, but a revolution. In like manner, in geological time, the tendency seems to be ever to disintegration and decay. This we see everywhere, and find that elevation occurs only by the introduction of
  • 71.
    new species ina way which is not obvious, and which may rather imply the intervention of a cause from without; so that here also we are required to admit as a general principle what is contrary to experience. If, however, we grant the evolutionist these postulates, we must next allow him to take the facts of botany and zoology out of their ordinary connection, and thread them like a string of beads, as Herbert Spencer has done in his “Biology,” on the threefold cord thus fashioned. This done, we next find, as might have been expected, certain gaps or breaks which require to be cunningly filled with artificial material, in order to give an appearance of continuity to the whole. The first of these gaps which we notice is that between dead and living matter. It is easy to fill this with such a term as protoplasm, which includes matter both dead and living, and so to ignore this distinction; but practically we do not yet know as a possible thing the elevation of matter, without the agency of a previous living organism, from that plane in which it is subject merely to physical force, and is unorganised, to that where it becomes organised, and lives. Under that strange hypothesis of the origin of life from meteors, with which Sir William Thomson closed his address at a late meeting of the British Association, there was concealed a cutting sarcasm which the evolutionists felt. It reminded them that the men who evolve all things from physical forces do not yet know how these forces can produce the phenomena of life even in its humblest forms. It is true that the scientific world has been again and again startled by the announcement of the production of some of the lowest forms of life, either from dead organic matter, or from merely mineral substances; but in every case heretofore the effort has proved as vain as the analogies attempted to be set up between the formation of crystals and that of organized tissues are fallacious. A second gap is that which separates vegetable and animal life. These are necessarily the converse of each other, the one deoxidizes and accumulates, the other oxidizes and expends. Only in reproduction or decay does the plant simulate the action of the animal, and the animal never in its simplest forms assumes the functions of the plant. Those obscure cases in the humbler spheres of animal and vegetable
  • 72.
    life which havebeen supposed to show a union of the two kingdoms, disappear on investigation. This gap can, I believe, be filled up only by an appeal to our ignorance. There may be, or may have been, some simple creature unknown to us, on the extreme verge of the plant kingdom, that was capable of passing the limit and becoming an animal. But no proof of this exists. It is true that the primitive germs of many kinds of humble plants and animal s are so much alike, that much confusion has arisen in tracing their development. It is also true that some of these creatures can subsist under very dissimilar conditions, and in very diverse states, and that under the specious name of Biology,[BA] we sometimes find a mass of these confusions, inaccurate observations and varietal differences made to do duty for scientific facts. But all this does not invalidate the grand primary distinction between the animal and the plant, which should be thoroughly taught and illustrated to all young naturalists, as one of the best antidotes to the fallacies of the evolutionist school. [BA] It is doubtful whether men who deny the existence of vital force have a right to call their science “Biology,” any more than atheists have to call their doctrine “Theology;” and it is certain that the assumption of a science of Biology as distinct from Phytology and Zoology, or including both, is of the nature of a “pious fraud” on the part of the more enlightened evolutionists. The objections stated in the text, to what have been called Archebiosis and Heterogenesis seem perfectly applicable, in so far as I can judge from a friendly review by Wallace, to the mass of heterogeneous material accumulated by Dr. Bastian in his recent volumes. The conclusions of this writer, would also, if established, involve evolution in a fatal embarras des richesses, by the hourly production during all geological time, of millions of new forms all capable of indefinite development. A third is that between any species of animal or plant and any other species. It was this gap, and this only, which Darwin undertook to fill up by his great work on the origin of species, but, notwithstanding the immense amount of material thus expended, it yawns as wide as ever, since it must be admitted that no case has been ascertained in which an individual of one species has transgressed the limits between it and other species. However extensive the varieties produced by artificial breeding, the essential characters of the species remain, and even its
  • 73.
    minor characters maybe reproduced, while the barriers established in nature between species by the laws of their reproduction, seem to be absolute. With regard to species, however, it must be observed that naturalists are not agreed as to what constitutes a species. Many so- called species are probably races, or varieties, and one benefit of these inquiries has been to direct attention to the proper discrimination of species from varieties among animals and plants. The loose discrimination of species, and the tendency to multiply names, have done much to promote evolutionist views; but the researches of the evolutionists themselves have shown that we must abandon transmutation of true species as a thing of the present; and if we imagine it to have occurred, must refer it to the past. Another gap is that between the nature of the animal and the self- conscious, reasoning, moral nature of man. We not only have no proof that any animal can, by any force in itself, or by any merely physical influences from without, rise to such a condition; but the thing is in the highest degree improbable. It is easy to affirm, with the grosser materialists, that thought is a secretion of brain, as bile is of the liver; but a moment’s thought shows that no real analogy obtains between the cases. We may vaguely suppose, with Darwin, that the continual exercise of such powers as animals possess, may have developed those of man. But our experience of animals shows that their intelligence differs essentially from that of man, being a closed circle ever returning into itself, while that of man is progressive, inventive, and accumulative, and can no more be correlated with that of the animal than the vital phenomena of the animal with those of the plant. Nor can the gap between the higher religious and moral sentiments of man, and the instinctive affections of the brutes, be filled up with that miserable ape imagined by Lubbock, which, crossed in love, or pining with cold and hunger, conceived, for the first time in its poor addled pate, “the dread of evil to come,” and so became the father of theology. This conception, which Darwin gravely adopts, would be most ludicrous, but for the frightful picture which it gives of the aspect in which religion appears to the mind of the evolutionist.
  • 74.
    The reader willnow readily perceive that the simplicity and completeness of the evolutionist theory entirely disappear when we consider the unproved assumptions on which it is based, and its failure to connect with each other some of the most important facts in nature: that, in short, it is not in any true sense a philosophy, but merely an arbitrary arrangement of facts in accordance with a number of unproved hypotheses. Such philosophies, “falsely so called,” have existed ever since man began to reason on nature, and this last of them is one of the weakest and most pernicious of the whole. Let the reader take up either of Darwin’s great books, or Spencer’s “Biology,” and merely ask himself as he reads each paragraph, “What is assumed here and what is proved?” and he will find the whole fabric melt away like a vision. He will find, however, one difference between these writers. Darwin always states facts carefully and accurately, and when he comes to a difficulty tries to meet it fairly. Spencer often exaggerates or extenuates with reference to his facts, and uses the arts of the dialectician where argument fails. Many naturalists who should know better are puzzled with the great array of facts presented by evolutionists; and while their better judgment causes them to doubt as to the possibility of the structures which they study being produced by such blind and material processes, are forced to admit that there must surely be something in a theory so confidently asserted, supported by so great names, and by such an imposing array of relations which it can explain. They would be relieved from their weak concessions were they to study carefully a few of the instances adduced, and to consider how easy it is by a little ingenuity to group undoubted facts around a false theory. I could wish to present here illustrations of this, which abound in every part of the works I have referred to, but space will not permit. One or two must suffice. The first may be taken from one of the strong points often dwelt on by Spencer in his “Biology.”[BB] [BB] “Principles of Biology,” § 118. But the experiences which most clearly illustrate to us the process of general evolution are our experiences of special evolution, repeated in every plant and animal. Each organism exhibits, within a short space
  • 75.
    of time, aseries of changes which, when supposed to occupy a period indefinitely great and to go on in various ways instead of one, may give us a tolerably clear conception of organic evolution in general. In an individual development we have compressed into a comparatively infinitesimal space a series of metamorphoses equally vast with those which the hypothesis of evolution assumes to have taken place during those unmeasurable epochs that the earth’s crust tells us of. A tree differs from a seed immeasurably in every respect—in bulk, in structure, in colour, in form, in specific gravity, in chemical composition: differs so greatly that no visible resemblance of any kind can be pointed out between them. Yet is the one changed in the course of a few years into the other; changed so gradually that at no moment can it be said, ‘Now the seed ceases to be and the tree exists.’ What can be more widely contrasted than a newly-born child and the small gelatinous spherule constituting the human ovum? The infant is so complex in structure that a cyclopædia is needed to describe its constituent parts. The germinal vesicle is so simple that it may be defined in a line.... If a single cell under appropriate conditions becomes a man in the space of a few years, there can surely be no difficulty in understanding how, under appropriate conditions, a cell may in the course of untold millions of years give origin to the human race. “It is true that many minds are so unfurnished with those experiences of nature, out of which this conception is built, that they find difficulty in forming it.... To such the hypothesis that by any series of changes a protozoan should ever give origin to a mammal seems grotesque—as grotesque as did Galileo’s assertion of the earth’s movement seem to the Aristoteleans; or as grotesque as the assertion of the earth’s sphericity seems now to the New Zealanders.” I quote the above as a specimen of evolutionist reasoning from the hand of a master, and as referring to one of the corner-stones of this strange philosophy. I may remark with respect to it, in the first place, that it assumes those “conditions” of evolution to which I have already referred. In the second place, it is full of inaccurate statements of fact, all in a direction tending to favour the hypothesis. For example, a tree
  • 76.
    does not differ“immeasurably” from a seed, especially if the seed is of the same species of tree, for the principal parts of the tree and its principal chemical constituents already exist and can be detected in the seed, and unless it were so, the development of the tree from the seed could not take place. Besides, the seed itself is not a thing self-existent or fortuitous. The production of a seed without a previous tree of the same kind is quite as difficult to suppose as the production of a tree without a previous seed containing its living embryo. In the third place, the whole argument is one of analogy. The germ becomes a mature animal, passing through many intermediate stages, therefore the animal may have descended from some creature which when mature was as simple as the germ. The value of such an analogy depends altogether on the similarity of the “conditions” which, in such a case, are really the efficient causes at work. The germ of a mammal becomes developed by the nourishment supplied from the system of a parent, which itself produced the germ, and into whose likeness the young animal is destined to grow. These are the “appropriate conditions” of its development. But when our author assumes from this other “appropriate conditions,” by which an organism, which on the hypothesis is not a germ but a mature animal, shall be developed into the likeness, of something different from its parent, he oversteps the bounds of legitimate analogy. Further, the reproduction of the animal, as observed, is a closed series, beginning at the embryo and returning thither again; the evolution attempted to be established is a progressive series going on from one stage to another. A reproductive circle once established obeys certain definite laws, but its origin, or how it can leave its orbit and revolve in some other, we cannot explain without the introduction of some new efficient cause. The one term of the analogy is a revolution, and the other is an evolution. The revolution within the circle of the reproduction of the species gives no evidence that at some point the body will fly off at a tangent, and does not even inform us whether it is making progress in space. Even if it is so making progress, its orbit of revolution may remain the same. But it may be said the reproduction of the species is not in a circle but in a spiral. Within the limit of experience it is not so, since, however it may undulate, it always returns into itself. But supposing it to be a spiral, it may ascend or descend, or expand and contract; but this does not
  • 77.
    connect it withother similar spirals, the separate origin of which is to be separately accounted for. I have quoted the latter part of the passage because it is characteristic of evolutionists to decry the intelligence of those who differ from them. Now it is fair to admit that it requires some intelligence and some knowledge of nature to produce or even to understand such analogies as those of Mr. Spencer and his followers, but it is no less true that a deeper insight into the study of nature may not only enable us to understand these analogies, but to detect their fallacies. I am sorry to say, however, that at present the hypothesis of evolution is giving so strong a colouring to much of popular and even academic teaching, more especially in the easy and flippant conversion of the facts of embryology into instances of evolution on the plan of the above extract, that the Spencerians may not long have to complain of want of faith and appreciation on the part of the improved apes whom they are kind enough to instruct as to their lowly origin. The mention of “appropriate conditions” in the above extract reminds me of another fatal objection to evolution which its advocates continually overlook. An animal or plant advancing from maturity to the adult state is in every stage of its progress a complete and symmetrical organism, correlated in all its parts and adapted to surrounding conditions. Suppose it to become modified in any way, to ever so small an extent, the whole of these relations are disturbed. If the modification is internal and spontaneous, there is no guarantee that it will suit the vastly numerous external agencies to which the creature is subjected. If it is produced by agencies from without, there is no guarantee that it will accord with the internal relations of the parts modified. The probabilities are incalculably great against the occurrence of many such disturbances without the breaking up altogether of the nice adjustment of parts and conditions. This is no doubt one reason of the extinction of so many species in geological time, and also of the strong tendency of every species to spring back to its normal condition when in any way artificially caused to vary. It is also connected with the otherwise mysterious law of the constant transmission of all the characters of the parent.
  • 78.
    Spencer and Darwinoccasionally see this difficulty, though they habitually neglect it in their reasonings. Spencer even tries to turn one part of it to account as follows:— “Suppose the head of a mammal to become very much more weighty—what must be the indirect results? The muscles of the neck are put to greater exertions; and the vertebras have to bear additional tensions and pressures caused both by the increased weight of the head and the stronger contraction of muscles that support and move the head.” He goes on to say that the processes of the vertebrae will have augmented strains put upon them, the thoracic region and fore limbs will have to be enlarged, and even the hind limbs may require modification to facilitate locomotion. He concludes: “Any one who compares the outline of the bison with that of its congener, the ox, will clearly see how profoundly a heavier head affects the entire osseous and muscular system.” We need not stop to mention the usual inaccuracies as to facts in this paragraph, as, for example, the support of the head being attributed to muscles alone, without reference to the strong elastic ligament of the neck. We may first notice the assumption that an animal can acquire a head “very much more weighty” than that which it had before, a very improbable supposition, whether as a monstrous birth Dr as an effect of external conditions after birth. But suppose this to have occurred, and what is even less likely, that the very much heavier head is an advantage in some way, what guarantee can evolution give us that the number of other modifications required would take place simultaneously with this acquisition! It would be easy to show that this would depend on the concurrence of hundreds of other conditions within and without the animal, all of which must co- operate to produce the desired effect, if indeed they could produce this effect even by their conjoint action, a power which the writer, it will be observed, quietly assumes, as well as the probability of the initial change in the head. Finally, the naivete with which it is assumed that the bison and the ox are examples of such an evolution, would be refreshing in these artificial days, if instances of it did not occur in almost every page of the writings of evolutionists.
  • 79.
    It would onlyweary the reader to follow evolution any further into details, especially as my object in this chapter is to show that generally, and as a theory of nature and of man, it has no good foundation; but we should not leave the subject without noting precisely the derivation of man according to this theory; and for this purpose I may quote Darwin’s summary of his conclusions on the subject.[BC] [BC] “Descent of Man,” part ii., ch. 21. “Man,” says Mr. Darwin, “is descended from a hairy quadruped, furnished with a tail and pointed ears, probably arboreal in its habits, and an inhabitant of the Old World. This creature, if its whole structure had been examined by a naturalist, would have been classed amongst the quadrumana, as surely as would the common, and still more ancient, progenitor of the Old and New World monkeys. The quadrumana and all the higher mammals are probably derived from an ancient marsupial animal; and this, through a long line of diversified forms, either from some reptile-like or some amphibian-like creature, and this again from some fish-like animal. In the dim obscurity of the past we can see that the early progenitor of all the vertebrata must have been an aquatic animal, provided with branchiæ, with the two sexes united in the same individual, and with the most important organs of the body (such as the brain and heart) imperfectly developed. This animal seems to have been more like the larvæ of our existing marine Ascidians than any other form known.” The author of this passage, in condescension to our weakness of faith, takes us no further back than to an Ascidian, or “sea-squirt,” the resemblance, however, of which to a vertebrate animal is merely analogical, and, though a very curious case of analogy, altogether temporary and belonging to the young state of the creature, without affecting its adult state or its real affinities with other mollusks. In order, however, to get the Ascidian itself, he must assume all the “conditions” already referred to in the previous part of this article, and fill most of the gaps. He has, however, in the “Origin of Species” and “Descent of Man,” attempted merely to fill one of the breaks in the evolutionary series, that between distinct species, leaving us to receive
  • 80.
    all the reston mere faith. Even in respect to the question of species, in all the long chain between the Ascidian and the man, he has not certainly established one link; and in the very last change, that from the ape-like ancestor, he equally fails to satisfy us as to matters so trivial as the loss of the hair, which, on the hypothesis, clothed the pre- human back, and on matters so weighty as the dawn of human reason and conscience. We thus see that evolution as an hypothesis has no basis in experience or in scientific fact, and that its imagined series of transmutations has breaks which cannot be filled. We have now to consider how it stands with the belief that man has been created by a higher power. Against this supposition the evolutionists try to create a prejudice in two ways. First, they maintain with Herbert Spencer that the hypothesis of creation is inconceivable, or, as they say, “unthinkable;” an assertion which, when examined, proves to mean only that we do not know perfectly the details of such an operation, an objection equally fatal to the origin either of matter or life, on the hypothesis of evolution. Secondly, they always refer to creation as if it must be a special miracle, in the sense of a contravention of or departure from ordinary natural laws; but this is an assumption utterly without proof, since creation may be as much according to law as evolution, though in either case the precise laws involved may be very imperfectly known. How absurd, they say, to imagine an animal created at once, fully formed, by a special miracle, instead of supposing it to be slowly elaborated through, countless ages of evolution. To Darwin the doctrine of creation is but “a curious illustration of the blindness of preconceived opinion.” “These authors,” he says, “seem no more startled at a miraculous act of creation than at an ordinary birth; but do they really believe that at innumerable periods in the earth’s history, certain elemental atoms have been commanded suddenly to flash into living tissues?” Darwin, with all his philosophic fairness, sometimes becomes almost Spencerian in his looseness of expression; and in the above extract, the terms “miraculous,” “innumerable,” “elemental atoms,” “suddenly,” and “flash,” all express ideas in no respect
  • 81.
    necessary to thework of creation. Those who have no faith in evolution as a cause of the production of species, may well ask in return how the evolutionist can prove that creation must be instantaneous, that it must follow no law, that it must produce an animal fully formed, that it must be miraculous. In short, it is a portion of the policy of evolutionists to endeavour to tie down their opponents to a purely gratuitous and ignorant view of creation, and then to attack them in that position. What, then, is the actual statement of the theory of creation as it may be held by a modern man of science? Simply this; that all things have been produced by the Supreme Creative Will, acting either directly or through the agency of the forces and materials of His own production. This theory does not necessarily affirm that creation is miraculous, in the sense of being contrary to or subversive of law; law and order are as applicable to creation as to any other process. It does not contradict the idea of successive creations. There is no necessity that the process should be instantaneous and without progression. It does not imply that all kinds of creation are alike. There may be higher and lower kinds. It does not exclude the idea of similarity or dissimilarity of plan and function as to the products of creation. Distinct products of creation may be either similar to each other in different degrees, or dissimilar. It does not even exclude evolution or derivation to a certain extent: anything once created may, if sufficiently flexible and elastic, be evolved or involved in various ways. Indeed, creation and derivation may, rightly understood, be complementary to each other. Created things, unless absolutely unchangeable, must be more or less modified by influences from within and from without, and derivation or evolution may account for certain subordinate changes of things already made. Man, for example, may be a product of creation, yet his creation may have been in perfect harmony with those laws of procedure which the Creator has set for His own operations. He may have been preceded by other creations of things more or less similar or dissimilar. He may have been created by the same processes with some or all of these, or by different means. His body may have been created in one way, his soul
  • 82.
    in another. Hemay, nay, in all probability would be, part of a plan of which some parts would approach very near to him in structure or functions. After his creation, spontaneous culture and outward circumstances may have moulded him into varieties, and given him many different kinds of speech and of habits. These points are so obvious to common sense that it would be quite unnecessary to insist on them, were they not habitually overlooked or misstated by evolutionists. The creation hypothesis is also free from some of the difficulties of evolution. It avoids the absurdity of an eternal progression from the less to the more complex. It provides in will, the only source of power actually known to us by ordinary experience, an intelligible origin of nature. It does not require us to contradict experience by supposing that there are no differences of kind or essence in things. It does not require us to assume, contrary to experience, an invariable tendency to differentiate and improve. It does not exact the bridging over of all gaps which may be found between the several grades of beings which exist or have existed. Why, then, are so many men of science disposed to ignore altogether this view of the matter? Mainly, I believe, because, from the training of many of them, they are absolutely ignorant of the subject, and from their habits of thought have come to regard physical force and the laws regulating it as the one power in nature, and to relegate all spiritual powers or forces, or, as they have been taught to regard them, “supernatural” things, to the domain of the “unknowable.” Perhaps some portion of the difficulty may be got over by abandoning altogether the word “supernatural,” which has been much misused, and by holding nature to represent the whole cosmos, and to include both the physical and the spiritual, both of them in the fullest sense subject to law, but each to the law of its own special nature. I have read somewhere a story of some ignorant orientals who were induced to keep a steam-engine supplied with water by the fiction that it contained a terrible djin, or demon, who, if allowed to become thirsty, would break out and destroy them all. Had they been enabled to discard this superstition, and to understand the force of steam, we can
  • 83.
    readily imagine thatthey would now suppose they knew the whole truth, and might believe that any one who taught them that the engine was a product of intelligent design, was only taking them back to the old doctrine of the thirsty demon of the boiler. This is, I think, at present, the mental condition of many scientists with reference to creation. Here we come to the first demand which the doctrine of creation makes on us by way of premises. In order that there may be creation there must be a primary Self-existent Spirit, whose will is supreme. The evolutionist cannot refuse to admit this on as good ground as that on which we hesitate to receive the postulates of his faith. It is no real objection to say that a God can be known to us only partially, and, with reference to His real essence, not at all; since, even if we admit this, it is no more than can be said of matter and force. I am not about here to repeat any of the ordinary arguments for the existence of a spiritual First Cause, and Creator of all things, but it may be proper to show that this assumption is not inconsistent with experience, or with the facts and principles of modern science. The statement which I would make on this point shall be in the words of a very old writer, not so well known as he should be to many who talk volubly enough about antagonisms between science and Christianity: “That which is known of God is manifest in them (in men), for God manifested it unto them. For since the creation of the world His invisible things, even His eternal power and divinity are plainly seen, being perceived by means of things that are made.”[BD] The statement here is very precise. Certain things relating to God are manifest within men’s minds, and are proved by the evidence of His works; these properties of God thus manifested being specially His power or control of all forces, and His divinity or possession of a nature higher than ours. The argument of the writer is that all heathens know this; and, as a matter of fact, I believe it must be admitted even by those most sceptical on such points, that some notion of a divinity has been derived from nature by men of all nations and tribes, if we except, perhaps, a few enlightened positivists of this nineteenth century whom excess of light has made blind. “If the light that is in man be darkness,
  • 84.
    how great isthat darkness.” But then this notion of a God is a very old and primitive one, and Spencer takes care to inform us that “first thoughts are either wholly out of harmony with things, or in very incomplete harmony with them,” and consequently that old beliefs and generally diffused notions are presumably wrong. [BD] Paul’s Epistle to the Romans, chap i. Is it true, however, that the modern knowledge of nature tends to rob it of a spiritual First Cause? One can conceive such a tendency, if all our advances in knowledge had tended more and more to identify force with matter in its grosser forms, and to remove more and more from our mental view those powers which are not material; but the very reverse of this is the case. Modern discovery has tended more and more to attach importance to certain universally diffused media which do not seem to be subject to the laws of ordinary matter, and to prove at once the Protean character and indestructibility of forces, the aggregate of which, as acting in the universe, gives us our nearest approach to the conception of physical omnipotence. This is what so many of our evolutionists mean when they indignantly disclaim materialism. They know that there is a boundless energy beyond mere matter, and of which matter seems the sport and toy. Could they conceive of this energy as the expression of a personal will, they would become theists. Man himself presents a microcosm of matter and force, raised to a higher plane than that of the merely chemical and physical. In him we find not merely that brain and nerve force which is common to him and lower animals, and which exhibits one of the most marvellous energies in nature, but we have the higher force of will and intellect, enabling him to read the secrets of nature, to seize and combine and utilize its laws like a god, and like a god to attain to the higher discernment of good and evil. Nay, more, this power which resides within man rules with omnipotent energy the material organism, driving its nerve forces until cells and fibres are worn out and destroyed, taxing muscles and tendons till they break, impelling its slave the body even to that which will bring injury and death itself. Surely, what we thus see in man must be the image and likeness of the Great Spirit. We can escape from this
  • 85.
    conclusion only byone or other of two assumptions, either of which is rather to be called a play upon words than a scientific theory. We may, with a certain class of physicists and physiologists, confine our attention wholly to the fire and the steam, and overlook the engineer. We may assume that with protoplasm and animal electricity, for example, we can dispense with life, and not only with life but with spirit also. Yet he who regards vitality as an unmeaning word; and yet speaks of “living protoplasm,” and “dead protoplasm,” and affirms that between these two states, so different in their phenomena, no chemical or physical difference exists, is surely either laughing at us, or committing himself to what the Duke of Argyll calls a philosophical bull; and he who shows us that electrical discharges are concerned in muscular contraction, has just as much proved that there is no need of life or spirit, as the electrician who has explained the mysteries of the telegraph has shown that there can be no need of an operator. Or we may, turning to the opposite extreme, trust to the metaphysical fallacy of those who affirm that neither matter, nor force, nor spirit, need concern them, for that all are merely states of consciousness in ourselves. But what of the conscious self this self which thinks, and which is in relation with surroundings which it did not create, and which presumably did not create it? and what is the unknown third term which must have been the means of setting up these relations? Here again our blind guides involve us in an absolute self-contradiction. Thus we are thrown back on the grand old truth that man, heathen and savage, or Christian and scientific, opens his eyes on nature and reads therein both the physical and the spiritual, and in connection with both of these the power and divinity of an Almighty Creator. He may at first have many wrong views both of God and of His works, but as he penetrates further into the laws of matter and mind, he attains more just conceptions of their relations to the Great Centre and Source of all, and instead of being able to dispense with creation, he hopes to be able at length to understand its laws and methods. If unhappily he abandons this high ambition, and contents himself with mere matter and physical force, he cannot rise to the highest development either of science or philosophy.
  • 86.
    It may, however,be said that evolution may admit all this, and still be held as a scientific doctrine in connection with a modified belief in creation. The work of actual creation may have been limited to a few elementary types, and evolution may have done the rest. Evolutionists may still be theists. We have already seen that the doctrine, as carried out to its logical consequences, excludes creation and theism. It may, however, be shown that even in its more modified forms, and when held by men who maintain that they are not atheists, it is practically atheistic, because excluding the idea of plan and design, and resolving all things into the action of unintelligent forces. It is necessary to observe this, because it is the half-way evolutionism which professes to have a Creator somewhere behind it, that is most popular; though it is, if possible, more unphilosophical than that which professes to set out from absolute and eternal nonentity, or from self-existent star-dust containing all the possibilities of the universe. Absolute atheists recognise in Darwinism, for example, a philosophy which reduces all things to a “gradual summation of innumerable minute and accidental material operations,” and in this they are more logical than those who seek to reconcile evolution with design. Huxley, in his “lay sermons,” referring to Paley’s argument for design founded on the structure of a watch, says that if the watch could be conceived to be a product of a less perfect structure improved by natural selection, it would then appear to be the “result of a method of trial and error worked by unintelligent agents, as likely as of the direct application of the means appropriate to that end, by an intelligent agent.” This is a bold and true assertion of the actual relation of even this modified evolution to rational and practical theism, which requires not merely this God “afar off,” who has set the stone of nature rolling and then turned His back upon it, but a present God, whose will is the law of nature, now as in times past. The evolutionist is really in a position of absolute antagonism to the idea of creation, even when held with all due allowance for the variations of created things within certain limits. Perhaps Paley’s old illustration of the watch, as applied by Huxley, may serve to show this as well as any other. If the imperfect watch,
  • 87.
    useless as atime-keeper, is the work of the contriver, and the perfection of it is the result of unintelligent agents working fortuitously, then it is clear that creation and design have a small and evanescent share in the construction of the fabric of nature. But is it really so? Can we attribute the perfection of the watch to “accidental material operations” any more than the first effort to produce such an instrument? Paley himself long ago met this view of the case, but his argument may be extended by the admissions and pleas of the evolutionists themselves. For example, the watch is altogether a mechanical thing, and this fact by no means implies that it could not be made by an intelligent and spiritual designer, yet this assumption that physical laws exclude creation and design turns up in almost every page of the evolutionists. Paley has well shown that if the watch contained within itself machinery for making other watches, this would not militate against his argument. It would be so if it could be proved that a piece of metal had spontaneously produced an imperfect watch, and this a more perfect one, and so on; but this is precisely what evolutionists still require to prove with respect both to the watch and to man. On the other hand it is no argument for the evolution of the watch that there may be different kinds of watches, some more and others less perfect, and that ruder forms may have preceded the more perfect. This is perfectly compatible with creation and design. Evolutionists, however, generally fail to make this distinction. Nor would it be any proof of the evolution of the watch to find that, as Spencer would say, it was in perfect harmony with its environment, as, for instance, that it kept time with the revolution of the earth, and contained contrivances to regulate its motion under different temperatures, unless it could be shown that the earth’s motion and the changes of temperature had been efficient causes of the motion and the adjustments of the watch; otherwise the argument would look altogether in the direction of design. Nor would it be fair to shut up the argument of design to the idea that the watch must have suddenly flashed into existence fully formed and in motion. It would be quite as much a creation if slowly and laboriously made by the hand of the artificer, or if more rapidly struck off by machinery; and if the latter, it would not follow that the machine which produced the watch was at all like the watch itself. It might have been something very different.
  • 88.
    Finally, when Spencertries to cut at the root of the whole of this argument, by affirming that man has no more right to reason from himself with regard to his Maker than a watch would have to reason from its own mechanical structure and affirm the like of its maker, he signally fails. If the watch had such power of reasoning, it would be more than mechanical, and would be intelligent like its maker; and in any case, if thus reasoning it came to the conclusion that it was a result of “accidental material operations,” it would be altogether mistaken. Nor would it be nearer the truth if it held that it was a product of spontaneous evolution from an imperfect and comparatively useless watch that had been made millions of years before. We have taken this illustration of the watch merely as given to us by Huxley, and without in the least seeking to overlook the distinction between a dead machine and a living organism; but the argument for creation and design is quite as strong in the case of the latter, so long as it cannot be proved by actual facts to be a product of derivation from a distinct species. This has not been proved either in the care of man or any other species; and so long as it has not, the theory of creation and design is infinitely more rational and scientific than that of evolution in any of its forms. But all this does not relieve us from the question, How can species be created?—the same question put to Paul by the sceptics of the first century with reference to the resurrection—“How are the dead raised, and with what bodies do they come?” I do not wish to evade this question, whether applied to man or to a microscopic animalcule, and I would answer it with the following statements:— 1. The advocate of creation is in this matter in no worse position than the evolutionist. This we have already shown, and I may refer here to the fact that Darwin himself assumes at least one primitive form of animal and plant life, and he is confessedly just as little able to imagine this one act of creation as any other that may be demanded of him. 2. We are not bound to believe that all groups of individual animals, which naturalists may call species, have been separate products of
  • 89.
    creation. Man himselfhas by some naturalists been divided into several species; but we may well be content to believe the creation of one primitive form, and the production of existing races by variation. Every zoologist and botanist who has studied any group of animals or plants with care, knows that there are numerous related forms passing into each other, which some naturalists might consider to be distinct species, but which it is certainly not necessary to regard as distinct products of creation. Every species is more or less variable, and this variability may be developed by different causes. Individuals exposed to unfavourable conditions will be stunted and depauperated; those in more favourable circumstances may be improved and enlarged. Important changes may thus take place without transgressing the limits of the species, or preventing a return to its typical forms; and the practice of confounding these more limited changes with the wider structural and physiological differences which separate true species is much to be deprecated. Animals which pass through metamorphoses, or which, are developed through the instrumentality of intermediate forms or “nurses”[BE] are not only liable to be separated by mistake into distinct species, but they may, tinder certain circumstances, attain to a premature maturity, or may be fixed for a time or permanently in an immature condition. Further, species, like individuals, probably have their infancy, maturity, and decay in geological time, and may present differences in these several stages. It is the remainder of true specific types left after all these sources of error are removed, that creation has to account for; and to arrive at this remainder, and to ascertain its nature and amount, will require a vast expenditure of skilful and conscientious labour. [BE] Mr. Mungo Ponton, in his book “The Beginning,” has based a theory of derivation on this peculiarity. 3. Since animals and plants have been introduced upon our earth in long succession throughout geologic time, and this in a somewhat regular manner, we have a right to assume that their introduction has been in accordance with a law or plan of creation, and that this may have included the co-operation of many efficient causes, and may have differed in its application to different cases. This is a very old doctrine of theology, for it appears in the early chapters of Genesis. There the
  • 90.
    Welcome to ourwebsite – the perfect destination for book lovers and knowledge seekers. We believe that every book holds a new world, offering opportunities for learning, discovery, and personal growth. That’s why we are dedicated to bringing you a diverse collection of books, ranging from classic literature and specialized publications to self-development guides and children's books. More than just a book-buying platform, we strive to be a bridge connecting you with timeless cultural and intellectual values. With an elegant, user-friendly interface and a smart search system, you can quickly find the books that best suit your interests. Additionally, our special promotions and home delivery services help you save time and fully enjoy the joy of reading. Join us on a journey of knowledge exploration, passion nurturing, and personal growth every day! ebookbell.com