Dravidian syntax and universal

grammar 1st Edition K.A. Jayaseelan

Visit to download the full and correct content document:
https://textbookfull.com/product/dravidian-syntax-and-universal-grammar-1st-edition-k
-a-jayaseelan/
More products digital (pdf, epub, mobi) instant
download maybe you interests ...

A Critical Account of English Syntax: Grammar, Meaning,

Text 1st Edition Keith Brown

https://textbookfull.com/product/a-critical-account-of-english-
syntax-grammar-meaning-text-1st-edition-keith-brown/

Den of Vipers 1st Edition K.A. Knight

https://textbookfull.com/product/den-of-vipers-1st-edition-k-a-
knight/

Biofueled Reciprocating Internal Combustion Engines 1st

Edition K.A. Subramanian

https://textbookfull.com/product/biofueled-reciprocating-
internal-combustion-engines-1st-edition-k-a-subramanian/

Basic Dutch: A Grammar and Workbook (Routledge Grammar

Workbooks) 2nd Edition Oosterhoff

https://textbookfull.com/product/basic-dutch-a-grammar-and-
workbook-routledge-grammar-workbooks-2nd-edition-oosterhoff/
English Syntax and Argumentation Fifth Edition Bas
Aarts

https://textbookfull.com/product/english-syntax-and-
argumentation-fifth-edition-bas-aarts/

Universal Chess Training 1st Edition Wojciech Moranda

https://textbookfull.com/product/universal-chess-training-1st-
edition-wojciech-moranda/

Grammar, Philosophy, and Logic 1st Edition Bruce Silver

(Auth.)

https://textbookfull.com/product/grammar-philosophy-and-
logic-1st-edition-bruce-silver-auth/

Software Languages Syntax Semantics and Metaprogramming

Ralf Lämmel

https://textbookfull.com/product/software-languages-syntax-
semantics-and-metaprogramming-ralf-lammel/

Java quick syntax reference Second Edition Olsson

https://textbookfull.com/product/java-quick-syntax-reference-
second-edition-olsson/
i

Dravidian Syntax and Universal Grammar

ii
OXFORD STUDIES IN COMPARATIVE SYNTAX
Richard Kayne, General Editor
Movement and Silence
Richard S. Kayne
Restructuring and Functional Heads: The
Cartography of Syntactic Structures, Volume 4
Guglielmo Cinque
Scrambling, Remnant Movement and
Restructuring in West Germanic
Roland Hinterhölzl
The Syntax of Ellipsis: Evidence from Dutch Dialects
Jeroen van Craenenbroeck
Mapping the Left Periphery: The Cartography of
Syntactic Structures, Volume 5
Edited by Paola Benincà and Nicola Munaro
Mapping Spatial PPs: The Cartography of Syntactic
Structures, Volume 6
Edited by Guglielmo Cinque and Luigi Rizzi
The Grammar of Q: Q-​Particles, Wh-​Movement,
and Pied-​Piping
Seth Cable
Comparisons and Contrasts
Richard S. Kayne
Discourse-​Related Features and Functional Projections
Silvio Cruschina
Functional Heads: The Cartography of Syntactic
Structures, Volume 7
Edited by Laura Brugé, Anna Cardinaletti,
Giuliana Giusti, Nicola Munaro, Cecilia Poletto
Adverbial Clauses, Main Clause Phenomena and
Composition of the Left Periphery: The Cartography
of Syntactic Structures, Volume 8
Liliane Haegeman
Variation in Datives
Edited by Beatriz Fernández and Ricardo Etxepare
Locality
Edited by Ian Roberts and Enoch Aboh
Aspects of Split Ergativity
Jessica Coon
A Comparative Grammar of Borgomanerese
Christina Tortora
Cross-​Linguistic Studies of Imposters and
Pronominal Agreement
Edited by Chris Collins
Japanese Syntax in Comparative Perspective
Edited by Mamoru Saito
Micro-​Syntactic Variation in North American
English
Edited by Raffaella Zanuttini and Laurence
R. Horn
Functional Structure from Top to Toe: The
Cartography of Syntactic Structures, Volume 9
Edited by Peter Svenonius
Chinese Syntax in a Cross-​linguistic Perspective
Edited by Edited by Y.-​H. Audrey Li,
Andrew Simpson, and W.-​T. Dylan Tsai
The Architecture of Determiners
Thomas Leu
Beyond Functional Sequence: The Cartography of
Syntactic Structures, Volume 10
Edited by Ur Shlonsky
The Cartography of Chinese Syntax: The Cartography
of Syntactic Structures, Volume 11
Edited by Wei-​Tien Dylan Tsai
Argument Licensing and Agreement
Claire Halpert
Rethinking Parameters
Edited by Luis Eguren, Olga Fernández-​Soriano,
and Amaya Mendikoetxea
Deconstructing Ergativity: Two Types of Ergative
Languages and Their Features
Maria Polinsky
iii

Dravidian Syntax
and Universal Grammar

K. A. Jayaseelan
R. Amritavalli

1
iv

1
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 certain other countries.

Published in the United States of America by Oxford University Press

198 Madison Avenue, New York, NY 10016, United States of America.

© Oxford University Press 2017

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 license, or under terms agreed with the appropriate reproduction
rights organization. Inquiries 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.

Library of Congress Cataloging-in-Publication Data

Names: Jayaseelan, K. A., author. | Amritavalli, R., author.
Title: Dravidian syntax and universal grammar : Jayaseelan-Amritavalli papers /
K.A. Jayaseelan, R. Amritavalli.
Description: New York, NY : Oxford University Press, [2017] |
Series: Oxford Studies in Comparative Syntax
Identifiers: LCCN 2016025663 | ISBN 9780190630225 (cloth : alk. paper) |
ISBN 9780190630249 (updf ) | ISBN 9780190630256 (ebook)
Subjects: LCSH: Dravidian languages—Syntax. | Dravidian languages—Grammar,
Comparative. | Generative grammar.
Classification: LCC PL4604 .J29 2016 | DDC 494.8—dc23
LC record available at https://lccn.loc.gov/2016025663

9 8 7 6 5 4 3 2 1
Printed by Sheridan Books, Inc., United States of America
╇v

Contents

Preface ix
Acknowledgments xiii

I | Scrambling and Word Order

1. IP-Internal Topic and Focus Phrasesâ•„3
K. A. Jayaseelan

2. Question Words in Focus Positionsâ•„44

K. A. Jayaseelan

3. Scrambling in the Cleft Construction in Dravidian 71

K. A. Jayaseelan and R. Amritavalli

4. Stacking, Stranding, and Pied-╉Piping: A Proposal about Word Order╄93

K. A. Jayaseelan

II | The Syntax of questions and quantifiers

5. Questions and Question-╉Word Incorporating Quantifiers in Malayalam╄129
K. A. Jayaseelan

6. Question and Negative Polarity in the Disjunction Phraseâ•„162

R. Amritavalli

7. Comparative Morphology of Quantifiersâ•„180

K. A. Jayaseelan

8. Question Particles and Disjunctionâ•„208

K. A. Jayaseelan

v
vi

vi Contents
9. Nominal and Interrogative Complements in Kannada‌222
R. Amritavalli

10. Decomposing Coordination: The Two Operators of Coordination 239

K. A. Jayaseelan

III | Finiteness and Negation

11. Kannada Clause Structure‌255
R. Amritavalli

12. Some Developments in the Functional Architecture of the Kannada Clause‌275

R. Amritavalli

13. Finiteness and Negation in Dravidian 299

R. Amritavalli and K. A. Jayaseelan

14. The Acquisition of Negation in Tamil 336

R. Amritavalli and Deepti Ramadoss

15. Coordination, Relativization and Finiteness in Dravidian ‌365

K. A. Jayaseelan

16. Separating Tense and Finiteness: Anchoring in Dravidian‌388

R. Amritavalli

IV | Case and Argument Structure

17. The Genesis of Syntactic Categories and Parametric Variation 417
R. Amritavalli and K. A. Jayaseelan

18. The Possessor/​Experiencer Dative in Malayalam‌434

K. A. Jayaseelan

19. The Serial Verb Construction in Malayalam‌453

K. A. Jayaseelan

20. The Argument Structure of the Dative Construction‌478

K. A. Jayaseelan

21. Syntactic Categories and Lexical Argument Structure‌490

R. Amritavalli

22. Parts, Axial Parts and Next Parts in Kannada‌502

R. Amritavalli
vi

Contents vii
23. The Dative Case in the Malayalam Verb‌517
K. A. Jayaseelan

24. Rich Results‌ 540

R. Amritavalli

V | Anaphors and Pronouns

25. Anaphorization in Dravidian‌ 573
R. Amritavalli

26. Anaphors as Pronouns‌ 589

K. A. Jayaseelan

‌27. Blocking Effects and the Syntax of Malayalam Taan 641

K. A. Jayaseelan

28. Deixis in Pronouns and Noun Phrases 659

K. A. Jayaseelan and M. Hariprasad

Language Index 675

Name Index 677
Subject Index 682
vi
ix

Preface

The papers included in this volume are a selection from the work on Dravidian done by the
two authors over the last thirty years—​the earliest paper here is dated 1984.
A brief introduction to Dravidian may be useful to readers who are unfamiliar with this
group of languages. The Dravidian languages are spoken principally in southern India.
But there are a few isolated Dravidian languages in the sub-​Himalayan belt, and one has
been discovered in Pakistan. Altogether 26 languages have been counted. But the principal
Dravidian languages are Tamil, Telugu, Kannada and Malayalam. These four languages, each
with its own writing system and long literary tradition—​the Tamil literary tradition dates
back to 500 BCE or earlier—​currently have millions of speakers, and are sometimes referred
to as the “major” Dravidian languages. These are the languages that figure in this volume.
There are typological descriptions of these four languages, which the reader who wishes to
have an over-​all picture of any one of them may wish to consult: Krishnamurti and Gwynn
(1985) (Telugu), Lehmann (1989) (Tamil), Sridhar (1990) (Kannada), Asher and Kumari
(1997) (Malayalam). For information about the geographical spread of the Dravidian lan-
guages, a list of these languages, speaker statistics, and the proto-​history of Dravidian, the
most accessible source is Krishnamurti (2003). A “principled typology” of just the anaphoric
systems of the above-​mentioned four principal languages can be found in Lust et.al. (2000).
A useful bibliographical tool for the Dravidian scholar is Ramaiah (1994–​2005), a six volume
bibliography of Dravidian languages and linguistics.
The papers in this volume are grouped into sections under five thematic heads. We now
give a brief indication of the main concerns of these sections, reserving a more detailed dis-
cussion of them to the mini-​prefaces that we give at the beginning of each section.
A commonly noted typological feature of the Dravidian languages is that they are head-​
final but that they otherwise have free word order. We deal with free word order—​always
ix
x

x Preface
foregrounded in typological accounts of Dravidian—​in the papers in Section I of this
volume. But we do not go along with a claim which has sometimes been made, that these lan-
guages have no neutral word order. Instead, the emphasis is on some hitherto unnoticed se-
mantic properties of Dravidian ‘scrambling.’ These are then related to parallel facts regarding
non-​canonical word orders in some OV languages of Europe; and we propose—​as a result of
these cross-​linguistic comparisons—​what we hope is a deeper explanation of this phenom-
enon in terms of the universal functional architecture of the clause.
Dravidian morphology is agglutinative. But what is interesting is that it is remarkably
transparent: in the case of quantifiers and questions, for example, Dravidian morphology
enables us to see the component parts out of which the quantifier and question meanings
are put together. We exploit this property of these languages—​this ‘window’ into syntax—​to
investigate questions and quantifiers, but again from a universalist perspective. This is the
burden of the papers of Section II.
The discussion of finiteness in section III begins with the problem that negative sentences
in Kannada are—​to all appearances—​matrix nonfinite clauses, which nevertheless have
tense interpretations. We propose that verbal aspect expresses the tense interpretation of the
clause, while it is Mood that expresses finiteness in Dravidian. A comparison of negative
clauses in the four principal Dravidian languages suggests how the interplay of agreement,
negation and positive polarity in verbal forms yields the observed variations in their negative
clauses. It has been a traditional claim in Dravidian linguistics that the finiteness of a clause
is indicated not by the presence of tense, but rather by that of verb agreement. Our analysis
makes sense of the traditional claim by treating verb agreement as a reflex of indicative mood;
we also extend the explanation to Malayalam, the lone Dravidian language with no overt
verb agreement. The papers here go further than traditional positions by exploring finiteness
in clause types that do not express agreement, and tense interpretation in clauses that have
no tense morpheme.
The papers in Section IV comprise two different topics. One is conjunct verbs, which are
ubiquitous in Dravidian sentences. We treat them as serial verbs, a phenomenon which has
been studied in a great many languages of the world. A second topic is the so-​called “dative
subject construction,” an equally prominent feature of Dravidian syntax. Differently from
most other accounts, we focus on the nature of the predicate in this construction, which is a
nominal; and we relate this fact to the absence (or near absence) of the adjective as a lexical
category in Dravidian. This leads on to a proposal about how the categories of adjective and
verb arise in Universal Grammar.
Some of the earliest work done by the two authors was on anaphora; and some of it is
presented in Section V. Going against Binding Theory’s well-​known opposition of anaphors
and pronouns—​the claim that syntactic contexts that allow one do not allow the other—​we
try to show that anaphors are in fact a subclass of pronouns. (We are perhaps closer to the
position of traditional grammar in saying this.) The section ends with a paper that looks
at the Dravidian pronominal system per se: deixis is inherently a part of every pronoun in
Dravidian—​‘he’ is always either ‘this he’ or ‘that he’. But we draw a universalist conclusion
from this: all definite expressions in languages have (even if only covertly) a deictic element,
‘this’ or ‘that’.
The chapters of this volume were written as individual papers by one or the other of
the authors, or jointly. No rewriting has been done to make them consistent with each
xi

Preface xi
other—​although we believe they do have an overall consistency as regards the main ideas.
In each section, the papers are given in their chronological order, so the reader can trace the
change and development of analyses and proposals. There is, inevitably, a certain amount of
overlap of ideas between papers; but cutting out these redundancies would have required
rewriting which (as we said) we have not attempted. One of the purposes we intend this
volume to serve is to make available in one place research work that originally appeared in
places that are now not easily accessible.

References

Asher, R.E. and Kumari, T.C. 1997. Malayalam. New York: Routledge.
Krishnamurti, Bh. 2003. The Dravidian Languages. Cambridge: Cambridge University Press.
Krishnamurti, Bh. and Gwynn, J.P.L. 1985. A Grammar of Modern Telugu. Delhi: Oxford
University Press.
Lehmann, T. 1989. A Grammar of Modern Tamil. Pondicherry: Pondicherry Institute of Language
and Culture.
Lust, B., K. Wali, J. Gair and K.V. Subbarao, eds. 2000. Lexical Anaphors and Pronouns in Selected
South Asian Languages: A Principled Typology. New York: Mouton de Gruyter.
Ramaiah, L.S. 1994-​2005. An International Bibliography of Dravidian Languages and Linguistics.
Madras: T. R. Publications Private Ltd. [Vol. 1: General and Comparative Dravidian Languages
and Linguistics, 1994; Vol. 2: Tamil Language and Linguistics, 1995; Vol. 3: Telugu Language
and Linguistics, 1998; Vol. 5: Malayalam Language and Linguistics, co-​author: N. Rajasekharan
Nair, 2001; Vol. 4: Kannada Language and Linguistics, co-​author: C.R. Karisiddappa, 2003;
Vol. 6: Non-​Literary Dravidian Languages and Linguistics, co-​author: B. Ramakrishna Reddy,
2005.]
Sridhar, S.N. 1990. Kannada. New York: Routledge.
xi
╇xi

Acknowledgments

Chapter 1, “IP-╉Internal Topic and Focus Phrases” [ Jayaseelan], originally appeared in Studia
Linguistica 55:1 (2001), 39–╉75, and is reprinted with permission from Blackwell.
Chapter 2, “Question Words in Focus Positions” [ Jayaseelan], originally appeared in
Linguistic Variation Yearbook, Vol. 3, Pierre Pica & Johan Rooryck eds. (2003), pp. 69–╉99.
Amsterdam: John Benjamins. It is reprinted with the publisher’s permission.
Chapter 3, “Scrambling in the Cleft Construction in Dravidian” [ Jayaseelan and Amritavalli],
originally appeared in The Free Word Order Phenomenon: Its Syntactic Sources and Diversity,
Joachim Sabel & Mamoru Saito eds. (2005), pp. 137–161. New York: Mouton de Gruyter. It
is reprinted with the publisher’s permission.
Chapter 4, “Stacking, Stranding and Pied-╉ Piping: A Proposal about Word Order”
[ Jayaseelan], originally appeared in Syntax 13:4 (2010), 298–330 and is reprinted with per-
mission from Blackwell.
Chapter 5, “Questions and Question-╉ Word Incorporating Quantifiers in Malayalam”
[ Jayaseelan], originally appeared in Syntax 4:2 (2001), 63–93 and is reprinted with permis-
sion from Blackwell.
Chapter 6, “Question and Negative Polarity in the Disjunction Phrase” [Amritavalli], orig-
inally appeared in Syntax 6:1 (2003), 1–╉18 and is reprinted with permission from Blackwell.
Chapter 7, “Comparative Morphology of Quantifiers” [ Jayaseelan], originally appeared in
Lingua 121:2 (2011), 269–286 and is reprinted with permission from Elsevier.
Chapter 8, “Question Particles and Disjunction” [ Jayaseelan], originally appeared in Linguistic
Analysis 38:1–╉2 (2012), 35–51 and is reprinted with permission from Linguistics Analysis.
Chapter 9, “Nominal and Interrogative Complements in Kannada” [Amritavalli], orig-
inally appeared in Deep Insights, Broad Perspectives: Essays in Honor of Mamoru Saito,
Y. Miyamoto, D. Takahashi, H. Maki, M. Ochi, K. Sugisaki & A. Uchibori eds. (2013),
pp. 1–╉21. Tokyo: Kaitakusha. It is reprinted with the publisher’s permission.

xiii
xvi

xiv Acknowledgments
Chapter 10, “Decomposing Coordination: The Two Operators of Coordination” [Jayaseelan],
originally appeared in Linguistic Analysis 40: 3–4 (2016), 237–253, and is reprinted with permis-
sion from Linguistic Analysis.
Chapter 11, “Kannada Clause Structure” [Amritavalli], originally appeared in The Yearbook
of South Asian Languages and Linguistics 2000, Rajendra Singh ed. (2000), pp. 11–30. New
Delhi: Sage India.
Chapter 12, “Some Developments in the Functional Architecture of the Kannada Clause”
[Amritavalli], originally appeared in Clause Structure in South Asian Languages, Veneeta
Dayal & Anoop Mahajan eds. (2004), 13–38. Boston: Kluwer Academic Publishers. It is
reprinted here with the publisher’s permission.
Chapter 13, “Finiteness and Negation in Dravidian” [Amritavalli and Jayaseelan], originally
appeared in The Oxford Handbook of Comparative Syntax, Guglielmo Cinque & Richard
S. Kayne eds. (2005), pp. 178–​220. New York: Oxford University Press. It is reprinted here
with the publisher’s permission.
Chapter 14, “The Acquisition of Negation in Tamil” [Amritavalli and Deepti Ramadoss],
originally appeared in Nanzan Linguistics Special Issue 1: Papers from the Consortium
Workshops on Linguistic Theory (2007), pp. 67–84. Graduate Program in Linguistic Science:
Nanzan University. We are happy that Deepti Ramadoss has given consent to this reprinting.
Chapter 15, “Coordination, Relativization and Finiteness in Dravidian” [ Jayaseelan], origi-
nally appeared in Natural Language and Linguistic Theory 32 (2014):191–211 and is reprinted
with permission from Springer.
Chapter 16, “Separating Tense and Finiteness: Anchoring in Dravidian” [Amritavalli],
originally appeared in Natural Language and Linguistic Theory 32 (2014): 283–306 and is
reprinted with permission from Springer.
Chapter 17, “The Genesis of Syntactic Categories and Parametric Variation” [Amritavalli and
Jayaseelan], originally appeared in Proceedings of the 4th GLOW in Asia 2003: Generative
Grammar in a Broader Perspective, Hang-​Jin Yoon ed. (2003), pp. 19–​41. The Korean
Generative Grammar Circle.
Chapter 18, “The Possessor/​Experiencer Dative in Malayalam” [ Jayaseelan], originally
appeared in Non-​nominative Subjects, Vol. 1, Peri Bhaskararao & Karumuri Venkata Subbarao
eds. (2004), pp. 227–​244. Amsterdam: John Benjamins. It is reprinted here with the pub-
lisher’s permission.
Chapter 19, “The Serial Verb Construction in Malayalam” [ Jayaseelan], originally appeared
in Clause Structure in South Asian Languages, Veneeta Dayal & Anoop Mahajan eds. (2004),
pp. 67–​91. Dordrecht:Kluwer Academic Publishers. It is reprinted here with the publisher’s
permission.
Chapter 20, “The Argument Structure of the Dative Construction” [ Jayaseelan], originally
appeared in Argument Structure, Eric Reuland, Tanmoy Bhattacharya and Giorgos Spathas
eds. (2007), pp. 37–48. Amsterdam: John Benjamins. It is reprinted here with the publisher’s
permission.
xv

Acknowledgments xv
Chapter 21, “Syntactic Categories and Lexical Argument Structure” [Amritavalli], originally
appeared in Argument Structure, Eric Reuland, Tanmoy Bhattacharya and Giorgos Spathas
eds. (2007), pp. 49–61. Amsterdam: John Benjamins. It is reprinted here with the publisher’s
permission.
Chapter 22, “Parts, Axial Parts and Next Parts in Kannada” [Amritavalli], originally appeared
in Tromso Working Papers on Language & Linguistics: Nordlyd 34.2, special issue on Space and
Scalar Structure, Monika Basic, Marina Pantcheva, Minjeong Son, and Peter Svenonius eds.
(2007), 1–​16. CASTL, Tromso.
Chapter 23, “The Dative Case in the Malayalam Verb” [ Jayaseelan], originally appeared
in Deep Insights, Broad Perspectives: Essays in Honor of Mamoru Saito, Y. Miyamoto,
D. Takahashi, H. Maki, M. Ochi, K. Sugisaki & A. Uchibori eds. (2013), pp. 139–​166.
Tokyo: Kaitakusha. It is reprinted here with the publisher’s permission.
Chapter 24, “Rich Results” [Amritavalli], originally appeared in The Lexicon-​Syntax inter-
face: perspectives from South Asian Languages, Pritha Chandra and Richa Srishti eds. (2014),
pp. 71–​100. Amsterdam: John Benjamins. It is reprinted here with the publisher’s permission.
Chapter 25, “Anaphorization in Dravidian” [Amritavalli], originally appeared in CIEFL
Working Papers in Linguistics 1.1 (1984), 1–31. Hyderabad: CIEFL.
Chapter 26, “Anaphors as Pronouns” [ Jayaseelan], originally appeared in Studia Linguistica
51:2 (1997), 186–234, and is reprinted with permission from Blackwell.
Chapter 27, “Blocking Effects and the Syntax of Malayalam taan” [ Jayaseelan], originally
appeared in The Yearbook of South Asian Languages and Linguistics 1998, Rajendra Singh ed.
(1998), pp. 11–​27. New Delhi: Sage Publications.
Chapter 28, “Deixis in Pronouns and Noun Phrases” [ Jayaseelan & M. Hariprasad], origi-
nally appeared in Linguistic Analysis 31:1–​2 (2001), 132–149 and is reprinted with permis-
sion from Linguistics Analysis. We are happy that M. Hariprasad has given consent to this
reprinting.
xvi
1

I Scrambling and Word Order

Preface

The papers in this section deal—​inter alia—​with Dravidian word order. The clausal order is
traditionally described as V-​final but with “free” order of the phrases in the pre-​verbal space.
The free order is traditionally generated by a rule of ‘scrambling’ in generative treatments of
Dravidian. But the proposal here is that the free order is capable of a more insightful expla-
nation. First of all, we need to understand the puzzle of rightward and leftward scrambling
in the pre-​verbal space. Rightward scrambling is to a position immediately to the left of V,
and it gives rise to focusing effects. This is the position of wh-​phrases in Malayalam. Leftward
scrambling has topicalizing effects. The interesting thing is that parallel scrambling patterns
have been noted in the OV languages of Europe. Our proposal—​which is in terms of clausal
architecture, i.e. the universal functional sequence (‘fseq’) of the clause—​is that there are IP-​
internal topic and focus phrases. This proposal explains scrambling patterns in Malayalam,
German, Dutch and Yiddish. See Chapters 1, 2, and 3.
Chapter 4 puts forward a general theory about head-​initial and head-​final word orders in
the world’s languages.
The literature on this general area is too vast for us to survey here. But the reader can
pursue the references in these papers for some guidance. The introductory essay of Sabel &
Saito (2005) gives a good conspectus of the issues relating to the topic of scrambling. Very
briefly, two broad types of scrambling have been postulated, namely the Japanese type which
apparently has no semantic consequences, and the type illustrated by the OV languages of
Europe (e.g. German and Dutch) which clearly has topicalizing or focusing effects. The claim
of our papers here is that Dravidian scrambling belongs to the second type. But this was not
the position of early generative work on Dravidian scrambling: see (e.g.) Mohanan (1982)
which makes the extreme claim that except for verb finality, the Malayalam sentence has no
neutral phrasal order. (This claim was made in the context of the early GB idea that there is
a fundamental divide between ‘configurational’ and ‘non-​configurational’ languages—​only
2

2 Scrambling and Word Order

the former have VP in their clause structure, while the latter have a “flat” clause structure. The
reader may see Speas (1990) for a critical discussion of this question.)
Regarding Chapter 4, which is about head-​initial and head-​final word orders, a good
point of entry to the issues here might perhaps be the papers in Svenonius (2000).
[The discussion of word order in these papers makes reference to the theory of ‘anti-
symmetry’ (Kayne 1994). A reader unfamiliar with this theory can find an outline of it in
Hornstein, Nunes & Grohmann (2005), ­chapter 7.]

References

Hornstein, N., J. Nunes & K. Grohmann. 2005. Understanding Minimalism. Cambridge:

Cambridge University Press.
Kayne, R. 1994. The Antisymmetry of Syntax. Cambridge, Mass.: The MIT Press.
Mohanan, K. P. 1982. Grammatical relations and clause structure in Malayalam. In J. Bresnan, ed.
The Mental Representation of Grammatical Relations, 504–​589. Cambridge, Mass.: The MIT
Press.
Sabel, J. & M. Saito (eds.). 2005. The Free Word Order Phenomenon: Its Syntactic Sources and
Diversity. New York: Mouton de Gruyter.
Speas, M. 1990. Phrase Structure in Natural Language. Dordrecht: Kluwer Academic Publishers.
Svenonius, P. (ed.) 2000. The derivation of VO and OV. Amsterdam: John Benjamins.
3

1 IP-​Internal Topic and Focus Phrases

K. A. Jayaseelan

In this paper I present a series of arguments for postulating a functional projection of

Focus above vP. I also postulate an iterable Topic Phrase above this Focus Phrase.
The postulation of IP-​internal Topic/​Focus projections will be shown to lead the way to
a new view of the difference between the clause structures of SOV and SVO languages, and
to some interesting results about clause-​internal scrambling and object shift in such diverse
languages as Malayalam, German, Dutch, Yiddish and Scandinavian.*

1. Question Words Contiguous to V

Many languages have a requirement that a question word should be contiguous to V.1 In
Malayalam, although the natural way to ask a question is by clefting, a non-​cleft question
is possible under one fairly strict condition: the question word must be placed immediately
to the left of V (in a position “normally” occupied by the direct object if one is present,
Malayalam being an SOV language):

(1) a. ninn-​e aarə aTiccu?

you-​acc. who beat(Past)
‘Who beat you?’
b. * aarə ninn-​e aTiccu?

(2) a. iwiTe aarə uNTə?

here who is
‘Who is here?’
b. * aarə iwiTe uNTə?
3
4

4 Scrambling and Word Order

(3) a. awan ewiTe pooyi?

he where went
‘Where did he go?’
b. * ewiTe awan pooyi?

(4) a. nii aa  pustakam aar-​kkə koDuttu?

you that book   who-​dat. gave
‘To whom did you give that book?’
b. * nii aar-​kkə aa pustakam koDuttu?

Even the clefting in questions, one can now see, is possibly a device for positioning the ques-
tion word next to V:

(5) nii entə aaNə tinn-​atə?

you what is ate-​Nominalizer
‘What did you eat?’ (Lit. ‘What is it that you ate?’)

In a cleft construction, the main verb is the copula; and the question word comes immedi-
ately to the left of the copula.2
How do we generate this position of the question word? Starting from an underlying SOV
word order of the type traditionally assumed in South Asian linguistics, it is difficult to see
how one can generate a COMP-​like position “within VP”. Equally impossible are the “down-
ward” movements we would need to postulate, to move (say) the subject into this position
(cf. (1a)); this is illustrated in (6):

(6) VP

SU V’

DO V’

? V

However, we can avoid these problems if we may assume a universal ‘Spec-​Head-​Complement’

order (Kayne 1994); and say that the surface order of the verb’s internal arguments in SOV
languages is the result of the raising of these arguments into SPECs of higher functional pro-
jections. While the subject raises to SPEC,IP, the internal arguments raise to SPECs of func-
tional projections which are intermediate between IP and VP. In the case of a monotransitive
verb (e.g.), the two movements shown in (7) would be the “normal” movements of an SOV
language. ((7) is anticipated, for Dutch, in Zwart (1993).)
5

IP-Internal Topic and Focus Phrases 5

(7) vP

SU v’

v VP

V DO

Given this picture, all we need to do, in order to generate the question word’s posi-
tion next to V, is to postulate a Focus Phrase (FP) immediately dominating vP, and to
say that the Q-​word moves into the SPEC of this FP. All other arguments (and such
adjuncts as are generated within vP, e.g. manner, location, time adverbials) would now
move “past” this position into SPECs of higher functional projections by the normal
movements which derive the SOV word order. In the case of (1a), e.g., the subject is
a Q-​word and moves into SPEC,FP and the direct object moves “past” it, as shown
in (8):

(8) FP

SPEC F’

F vP

SU v’

v VP

V DO

(V adjoins to v; there is reason to think that [v V-​v] adjoins to Focus.3)

(Let us note at this point that some proposals in the literature for functional pro-
jections “in the middle” of vP/​VP are irrelevant for our purposes. Koizumi’s “split VP”
hypothesis (Koizumi 1994) has an AGRoP “between” a higher VP in which the subject
is generated and a lower VP in which the internal arguments are generated; Collins &
Thrainsson (1996) add a TP immediately above this AGRoP:

(9) (Collins & Thrainsson 1996, p.401; irrelevant details omitted)

6

6 Scrambling and Word Order

VP1

SU V’

V1 TP

T AGRoP

AGRo VP2

V2 DO
    

Note however that these “intermediate” functional projections are lower than the subject’s
base position; even if we were to add (say) a CP immediately above the TP of (9), the subject
will have to be “lowered” into it to generate a sentence like (1a).)4

2. Scrambling in Malayalam and the Structure of the Malayalam Clause

In what is usually taken to be “the VP” of SOV languages, the canonical order of elements
is: Adjunct—​IO—​DO—​V; cf. (10):

(10) ñaan innale Mary-​k’k’ə oru kattə ayaccu

I yesterday   -​ dat. a letter sent
‘I sent a letter to Mary yesteday.’

As a comparison of this sentence with its English gloss shows, the order of elements is the
mirror-​image of English:

Adjunct -- IO -- DO -- V -- DO -- IO -- Adjunct

The movements out of the VP that we postulated for SOV languages are apparently “nested”
movements.5
Interesting questions arise about Relativized Minimality. How do these movements es-
cape minimality effects? There are two sub-​questions. One, if SPEC,FP is filled, how
do these movements go past it—​or (indeed) past SPEC, vP (the “VP-​internal subject”)?
Two, why are there no inter se minimality effects among them; e.g. why doesn’t the land-
ing site of the direct object prevent the indirect object moving to a higher position? The
problem (of course) is that the Malayalam V does not raise any higher than the head of FP
(as we just said).6
7

IP-Internal Topic and Focus Phrases 7

It has been recognized (however) that we need to postulate two types of movements: one,
instantiated by Icelandic object shift, which obeys minimality; the other, instantiated by
scrambling in Dutch, which does not obey this constraint (Zwart 1993, Diesing 1997).While
the reason for this distinction remains puzzling—​especially since both types of movement
have many things in common; e.g. they obey a common ‘definiteness/​specificity’ constraint -​,
let us for the time being simply say that the migration of arguments and adjuncts out of the
VP (in SOV languages) is a case of scrambling.
What are the functional heads, higher than FP and lower than IP, which host these moved
phrases? It has been claimed for the COMP system (Rizzi 1997), that there are any number
of Topic Phrases possible above the FP in COMP; assuming a similar possibility with respect
to the FP above vP, it is tempting to say that the “normal” movements of the internal argu-
ments (and adjuncts) of SOV languages are to SPEC,TopP.7
This solution has a seeming advantage: repeated applications of Topicalization should be
able to produce any order whatever of the elements that undergo the operation. I.e. the base
order of these elements can be arbitrarily reordered. This should be able to generate “scram-
bling” understood as the free order of a verb’s arguments (which is the “classical” view of
scrambling). This advantage however is outweighed by other considerations. Firstly, it can-
not account for the canonical order of the verb’s internal arguments, the one which we tried
to describe in terms of “nested” movements.
But the really serious problem is that the internal arguments in their canonical order—​
as, for example, in (10)—​do not show any topicalization effects. Topics are familiar in-
formation in the discourse; they are entities which have already been mentioned, and are
therefore definite or specific. (In fact, we shall be arguing that the leftward movements
showing a definiteness/​specificity effect in Scandinavian, Dutch or Yiddish are instances of
topicalization—​specifically, of movement into TopPs above FP.) But there are no definite-
ness/​specificity constraints on the Malayalam verb’s internal arguments in their canonical
order; cf.

(11) en-​ik’k’ə oru aana-​ye weeNam

I-​dat. an elephant-​acc. want
‘I want an elephant.’

(12) nii puuwə paRik’k’-​arutə

you flower pick-​should not
‘You should not pick flowers.’

(13) awan oru maNDan aaNə

he an idiot is
‘He is an idiot.’

It is difficult to imagine why the indefinite NPs in (11)-​(13) should be topicalized.

Our “nested” movements (then) are not into Topic Phrases.8 However, suppose we do postu-
late iterable TopPs above FP in Malayalam also—​on the evidence of the European languages, to
which we come back presently. Do we say that the “nested” movements that we are postulating
are into positions higher than TopP*, or lower than TopP*? I.e., is (14) or (15) the correct picture?
8

8 Scrambling and Word Order

(14) IP

SPEC I’

TopP*

Top FP

F vP

(15) IP

SPEC I’

I TopP*

FP

F vP

When two definite arguments interchange their positions, it is difficult to say which one
is topicalized; e.g.

(16) a. ñaan Mary-​k’k’ə aa pustakam koDuttu

I   -​ dat. that book gave
b. ñaan aa pustakam Mary-​ k’k’ə koDuttu
I that book -​dat. gave
‘I gave that book to Mary.’
9

IP-Internal Topic and Focus Phrases 9

(17) a. ñaan Mohanan-​ooDə aa kaaryam paRaññiTT-​illa

I     -​ 2nd dat. that matter told-​neg.
b. ñaan aa    kaaryam Mohanan-​ooDə paRaññiTT-​illa
I that matter      -​ 2nd dat. told-​neg.
‘I haven’t told Mohanan (about) that matter.’

The IO-​D O order of (16a)/​(17a) is the canonical order; the scrambled order DO-​IO of
(16b)/​(17b) could have been produced either by the IO moving (from vP) into a TopP
below the canonical position (assuming (14)),9 or by the DO moving (from vP) into
a TopP above the canonical position (assuming (15)). This can be schematically repre-
sented as (18). (The “unoccupied” canonical position in each case is indicated within
parentheses.)

(18) a. (IO) DO IOi . . . [vP . . . ti . . . ]

b. DOi IO (DO) . . . [vP . . . ti . . . ]

However, if one of the arguments is indefinite, we get some interesting results. Cf.

(19) a. ñaan awan-​ə oru kattə ayaccu

I he-​dat. a letter sent
b. ?* ñaan oru kattə awan-​ə ayaccu
I a  letter he-​ dat. sent
‘I sent him a letter.’

(20) a. ñaan awaL-​kkə paNam koDuttu

I she-​dat. money gave
b. ?* ñaan paNam awaL-​kkə koDuttu
I money she-​dat. gave
‘I gave her money.’

(21) a. ñaan ninn-​ooDə oru tamaas´a paRay-​aam

I you-​2nd dat. a joke say-​will
b. ?* ñaan oru tamaas´a ninn-​ooDə   paRay-​aam
I a joke    you-​2nd dat. say-​will
‘I’ll tell you a joke.’

If the interchange of positions in the (b) sentences is due to IO moving into a TopP below
its canonical position, it is difficult to see why these sentences are unacceptable, since a
10

10 Scrambling and Word Order

definite pronoun is always amenable to topicalization. On the other hand, if what is hap-
pening in the (b) sentences is the movement of DO into a TopP above its canonical posi-
tion, the ungrammaticality of these sentences is explained: an indefinite (nonspecific) NP
has been (illicitly) topicalized. These data (then) support (15) over (14).
As a matter of fact, if the IO is indefinite and the DO definite—​the reverse of what is the
case in (19)-​(21)—, the canonical order is somewhat awkward! This is especially so, if the DO
is a pronoun, cf.

(22) a. ?? ñaan oru bhikshakkaaran-​ə atə koDuttu

I a beggar-​dat. it gave
b. ñaan atə oru  bhikshakkaaran-​ə koDuttu
I it a    beggar-​
dat. gave
‘I gave it to a beggar.’

If what is happening in (b) is the IO being topicalized in a position lower than its
canonical position, the acceptability of this sentence is puzzling—​since an indefinite
NP is being topicalized. But if the (definite) DO is being topicalized in a position
higher than its canonical position, the complete acceptability of the (b) sentence is
unsurprising.10
A very interesting pair of sentences which helps us to choose between (15) and (14) is the
following:

(23) a. ñaan oru maratt-​inə weLLam ozhiccu

I a  tree-​
dat. water poured
‘I poured water to a tree.’
b. ñaan weLLam oru maratt-​inə ozhiccu
I water a tree-​dat. poured
‘I poured the water to a tree.’

In Malayalam, the definite article is null. This means that in itself, a form like weLLam
(‘water’) is ambiguous between a definite and an indefinite reading. In the (a) sentence,
which has the canonical order, the most natural interpretation of weLLam is as ‘(some)
water’; i.e. the argument is indefinite. But in the (b) sentence, which has the inverse order,
the only permissible interpretation of weLLam is as ‘the water’; i.e. the argument is ob-
ligatorily definite. (The (a) sentence could be an answer to the question ‘What did you
do?’ The (b) sentence could only be an answer to the question ‘What did you do with the
water?’) This definiteness constraint on weLLam in the (b) sentence is explained if it is a
Topic. I.e., in effect, we are choosing the structure shown in (15), and saying that the DO is
topicalized in a position higher than its canonical position. But if we were to choose (14)
and claim that it is the IO which is topicalized in this sentence, in a position lower than
its canonical position, then we end up with two problems: Why is there a definiteness
constraint on the DO, which we are now saying is in its canonical position? How can an
indefinite and non-​specific NP like oru maratt-​inə (‘one tree-​dat.’) be topicalized?11
1

IP-Internal Topic and Focus Phrases 11

We (then) choose (15) as correctly representing the configuration of TopPs with respect to
the canonical positions of arguments in SOV languages.12
However recall that in Rizzi’s (1995) articulation of the COMP system of IP, there are
TopPs both above and below FP (see fn. 7). Is there any evidence of TopPs below FP in the
COMP system of vP (also)? Consider the following sentences:

(24) aarum kaND-​illa, aana-​ye

nobody saw-​neg. elephant-​acc.
‘The elephant, nobody saw.’

(25) aarə ayaccu, ninn-​e ?

who sent you-​acc.
‘You, who sent?’

If we may assume a TopP below FP, we can readily explain the post-​verbal elements in these
sentences. We can say that these elements are in this TopP; and that (furthermore) aarum
‘nobody’ (a negative polarity item) in (24), and aarə ‘who’ (a question word) in (25) are in
SPEC,FP, and that V has raised and adjoined to F.
Tirumalesh (1996) was the first to claim (to my knowledge) that in Dravidian, the right-​
of-​V elements are Topics. He pointed out that “(non-​generic non-​human) indefinite noun
phrases” were unacceptable in this position. Thus consider (26):

(26) ?* ñaan awan-​ə ayaccu, oru kattə

I he-​dat. sent a letter

(26) is as unacceptable as (19b) (repeated below), or (27) (where the indefinite NP is in

a Topic position above IP):

(19b) ?* ñaan oru kattə awan-​ə ayaccu

I a letter he-​dat. sent

(27) ?* oru kattə ñaan awan-​ə ayaccu

a letter I he-​dat. sent13

Our postulation of a Topic position below FP (and therefore below the canonical
positions of the verb’s internal arguments) may appear to compromise our expla-
nation of sentences like (19)-​(21), in which a definite IO and an indefinite DO
interchange their linear order. (19b) has been repeated in the last paragraph. Our ex-
planation of its ungrammaticality was that the indefinite DO oru kattə ‘a letter’ has
been topicalized. But surely, the same word order could be obtained by moving the
definite IO awan-​ə ‘he-​dat.’ into the below-​FP Topic position? The fact of the matter
is that (for unclear reasons) the below-​FP Topic position is entirely ‘defocused’ and
seems to induce obligatory V-​raising past it. In fact, this position seems infelicitous
12

12 Scrambling and Word Order

if it does not occur in association with FP, to the head of which V raises and adjoins.
Cf. (28):

(28) a. ñaan aa pustakam waangi

I    that book bought
‘I bought that book.’
b. aa pustakam ñaan waangi
c. ?* ñaan waangi, aa pustakam
d. ñaan-​um waangi, aa pustakam
I-​conj. bought that book
‘I too bought that book.’

(28a) has the canonical word order. In (28b), aa pustakam ‘that book’ has been moved into a
pre-​subject Topic position, which is fine. The unacceptability of (28c) is due to the fact that
there is no F to induce V-​raising. In (28d), ñaan-​um ‘I too’ is in SPEC,FP and V adjoins to
F; and the sentence is fine.
However, the fact that the below-​FP Topic position invariably appears (in linear terms)
post-​verbally, suggests also another analysis of these data. Consider (29) ((29b)=(24)):

(29) a. aana-​ye aarum kaND-​illa

elephant-​acc. nobody saw-​neg.
‘The elephant, nobody saw.’
b. aarum  kaND-​ illa, aana-​ye
nobody saw-​neg.   elephant-​acc.

In the (a) sentence, aana-​ye ‘elephant-​acc.’ is plausibly in a pre-​IP Topic position. The (b) sen-
tence, we could suggest, is derived from the (a) sentence by preposing IP to the SPEC of a
still higher functional head (say, a higher TopP). To account for the marginal status of a sen-
tence like (28c), one could now say that IP-​preposing requires a focused element in the IP.14
In fact, the preposing of IP and VP must be assumed to take place quite generally in
Malayalam, as a result of the familiar property of SOV-​language verbs of moving their argu-
ments to the left (the “nested” movements to canonical positions that we spoke of )—​but
extended now (at least in Malayalam) to auxiliary verbs and the verbal complementizer. First,
note that the auxiliary verbs are stacked on the right-​hand side of the lexical verb, in the in-
verse order of English; e.g.

(30) awan atə tinn-​iTTuND-​aakaam

he it eat-​perf.-​may
‘He may have eaten it.’

The VPs headed by the auxiliary verbs, we may assume, are generated still higher than the
Topic and Focus positions—​and the “sandwiched” canonical positions—​that we investi-
gated. We can reasonably claim that the auxiliary Vs share with the main V the property
Another random document with
no related content on Scribd:
carried to a safe place and saying, “I will face the enemy.” If the
battle has its monument, so the hero that won it has his, and the
traveler on the New York Central Railway can see both, but thirty
miles apart, the one at Oriskany, the other a short distance down the
valley from Little Falls.

Fig. 10. Oriskany Battle Monument, a Few Miles West of

Utica
Herkimer was not a trained soldier, but a plain farmer of the
valley. His letters and military orders show us that he could spell as
poorly as any of his neighbors, and that is saying a good deal. His
army was made up of these same simple neighbors, who, though
they did not know much about soldierly marching, were good shots
and hard hitters, fighting not for pay but to save their liberty and to
protect their homes from the cruel savages.
The names of many of these men are on the battle monument,—
names such as Groot, Petrie, Dunckel, Klock, Kraus, Sammons,
Schnell, Van Horn, and Zimmerman. We do not need to be told that
these were not men of English blood; indeed, many of them
belonged to those same Dutch families which we saw settling in the
Hudson and Mohawk valleys. And some, like the last one, were not
Dutch but German, and their ancestors came not from Holland but
from a land farther up the Rhine. They had been driven out by the
persecutions of one of the French kings and had come to America.
They had had a hard time, suffering much from taskmasters, from
poverty, and from the savages, until finally they had gone farther
west in the Mohawk valley and had received good lands lying
eastward from Utica. There they became comfortable and
prosperous. They answered promptly the brave Herkimer’s call to
arms, and many of them gave their lives for home and country at
Oriskany.
We must now tell the other side of the story and see who the
invaders were and where they came from. In Revolutionary days
nearly all the people of New York were in its two great valleys. One
could go up the Hudson from New York, pass Albany and Fort
Edward, and, without finding high ground, enter the valley of lake
Champlain and go down to Montreal on the St. Lawrence. Here,
then, was an easy valley road from the sea at New York into
Canada. Coming either way, one could turn off to the west at Fort
Orange or Albany and go up the Mohawk and down to Oswego on
lake Ontario. In these two valleys were all the farms, the towns, and
of course the forts. There were forts at Oswego and where Rome,
Utica, and Albany are; at Fort Edward, Fort Ann, Ticonderoga, and
many other places, making a chain of defenses in these valleys.
West of the Hudson and south of the Mohawk were the high, rough
woods of the Catskills; while west of lake Champlain and north of the
Mohawk were the rugged Adirondacks, without roads or clearings.
And because the roads, the homes, and the forts were in the valleys,
we shall almost always find the armies and the fighting there.
This will help us to understand the plan which the British made in
1777, by which they felt sure of crushing the rebellion. The year
before they had to leave Boston and had come around to New York.
New York was not so large as Philadelphia then, but it was an
important place, for it was the key to the Hudson valley. The British
generals decided to send one army up the Hudson to destroy the
forts and beat back the colonists. This army was under General
Howe. Another army, commanded by General Burgoyne, was to
come from the St. Lawrence up lake Champlain and through the
woods by Fort Edward to Albany. Burgoyne was a brave officer, but
he was conceited, and he felt too sure that he could do his part
easily. He was confident that when he marched through the country
many colonists would run to place themselves under the English
flag. In a few weeks he learned that these backwoods Americans
were quite ready to meet and give battle to the combined forces of
the British regulars, the hired German soldiers, and the Indians with
whom they were in league.

Fig. 11. General Nicholas Herkimer directing the Battle of

Oriskany
 There was yet a third division in this campaign. A British force
under General St. Leger had come up the St. Lawrence and lake
Ontario to Oswego. St. Leger also had with him many Indians, and
these were commanded by Joseph Brant, a famous chief, who had
had much to do with white men and who was well educated. This
third army was to go east, over the Oneida Carrying Place and down
the Mohawk to Albany. By this pretty plan three armies, one from the
south under Howe, one from the north under Burgoyne, and one
from the west under St. Leger, were to meet in Albany. They would
put British soldiers in every fort on the way, capture and disarm the
rebels, and have all New York under their feet. More than this, they
would thus shut off New England from Pennsylvania and Virginia,
cutting the unruly colonies into two parts so that they could not help
each other.
But the scheme, brilliant as it was, would not work. None of the
British armies reached Albany. Howe did not, perhaps because he
did not try. Burgoyne and St. Leger did not, because they could not:
there was altogether too much in the way. We shall now see how this
happened.
St. Leger brought into the Mohawk valley from Oswego an army
of seventeen hundred men. Some were British, some were Hessians
or hired German soldiers, and the rest were Indians under Joseph
Brant. They thought that it would not be much trouble to take Fort
Stanwix and then go down the valley, burning and killing as they
went, until they should meet the other armies of the king at Albany.
But the colonists sent more soldiers to defend the fort, and Colonel
Peter Gansevoort, who was in command, had under him nearly a
thousand men. Just before the British came in sight a stock of
provisions, brought on several boats up the river, had been safely
delivered within the defenses. This was early in August, and only
about seven weeks before Congress had adopted the style of
American flag which we know so well. There was no flag at Fort
Stanwix, so the garrison set about making one. They cut up shirts to
make the white. The blue came from a cloak captured not long
before in a battle, on the Hudson, by Colonel Marinus Willett, one of
the bravest commanders within the fort. The red is said to have been
taken from a petticoat. Certain it is that a patriot flag was made, and
some think that it was the first American flag ever raised over a
fortification.
While the British were besieging Fort Stanwix, General Herkimer
had called out the men of the valley, bidding all between the ages of
sixteen and sixty make ready for battle. The boys and old men were
to do their best to care for the families and to defend their homes.
Eight hundred men gathered under Herkimer and marched to help
the garrison of the fort. Hearing of this, part of the British army,
including the Indians, came down the valley to head off Herkimer.
They met at Oriskany. The farmer soldiers were hurrying up the
valley without due watching for sudden attack, while the enemy
placed themselves in ambush around a low field which was wooded
and swampy. Through this field the road ran, and when Herkimer’s
men were well down into it the Indians opened a hot fire, which threw
the patriots into disorder. They soon rallied and fought fiercely for
five hours, until two hundred of them had lost their lives. Early in the
battle Herkimer was shot, but he forgot his pain when he saw his
men victorious. Much of the fighting was of the Indian sort, from
behind trees, for the Dutchmen well knew the ways of the savages.
They saw that when one man fired from behind a tree an Indian
would rush forward to tomahawk him before he could load his gun
for another shot. So they were ordered to stand by twos and take
turns in firing. Thus when the Indian ran forward with his tomahawk
he would receive a bullet from the other man’s gun.
 Fig. 12. Nicholas Herkimer’s Monument
To the right is the old mansion in which he lived. Near Little Falls, New York
Under John Johnson, the son of Sir William Johnson, were many
Tories from the valley. They and the patriots often recognized each
other as former neighbors, and then the fight was more stubborn
than ever, for the soldiers of freedom were bitterly angry to find old
friends in arms against them. During the battle a terrific thunder-
shower came up, and both sides stopped fighting, having enough to
do to keep their powder and guns dry. The dark storm passed and
the strife went on again. At length the Indians grew tired and ran,
leaving the field to Herkimer and his little army. The importance of a
conflict is not always in proportion to the size of the armies engaged,
and in what it did for freedom Oriskany takes high place among the
battles of modern times.
The enemy went back to the siege of Fort Stanwix, and soon a
new force of patriots under Benedict Arnold was sent up the valley to
relieve the fort. It was during this march that an ignorant but cunning
fellow named Han Yost Schuyler was caught, tried, and condemned
to die as a spy. Because his friends pleaded for his life Arnold finally
told him that he might live if he would go up to Fort Stanwix and
make the Indians and British believe that a great army was marching
against them. Meanwhile the man’s brother was held as a hostage,
to be punished if the promise was not fulfilled. Han Yost did his part
so well that St. Leger, taking fright, left the fort in great haste and his
expedition was entirely broken up. Why he did not have a gay march
down to Albany is now quite plain.
A few days after the battle of Oriskany a number of men drove
some cattle to Fort Stanwix as food for the soldiers. Several women
went with them on horseback to visit their husbands, who belonged
to the garrison. At the ford of the river, now the Genesee street
crossing in Utica, a big Dutchman, who did not wish to get wet,
leaped uninvited upon a horse behind one of the women. The horse
did not like the double load, and made great sport by throwing the
Dutchman into the middle of the stream, while he carried his
mistress over in safety.
General Burgoyne came nearer Albany than did St. Leger.
Indeed he went to Albany, but not until he had lost his army. He had
promptly captured Ticonderoga on lake Champlain, and this success
gave him high hopes and sent rejoicing throughout Great Britain; but
the patriots, by felling trees and cutting away bridges, hindered his
southward march in every way. He sent a thousand of his German
soldiers across to Bennington, among the Green mountains, to
capture stores which he knew were there. But General Stark was
there also, with a little army from New Hampshire and
Massachusetts, and the thousand Hessians did not go back to help
Burgoyne. He had left another thousand to guard Ticonderoga, and
so he was two thousand short. All this time the patriot army was
growing, for the men of the Hudson valley were maddened when
they saw the bloodthirsty Indians marching with the English, and, to
Burgoyne’s surprise, they had no mind to fight for the king. Howe did
not come, St. Leger did not come, and the provisions were getting
short. These could only come along the road from the north, and the
colonists were already marching in behind Burgoyne’s army to cut
his line of communications. He knew that he must fight or starve. He
chose to fight. The battle was fought on Bemis Heights, a range of
hills west of the Hudson, a short distance north of the little village of
Stillwater. The British general, after his defeat, withdrew a few miles
northward and surrendered his army near the present town of
Schuylerville. A tall monument marks the place. This was the battle
of Saratoga, fought in old Saratoga, which is several miles from the
famous resort of that name.
So it was that up and down these beautiful valleys went armies
and scouting bands, as well as peaceful emigrants with their oxen,
their stages, and their small freight boats. One cannot go far along
the Hudson or the Mohawk without finding the site of an Indian
village, the foundations of an old fort, the homestead of a
Revolutionary hero, or an ancient place of worship. When we see the
great railways and swift trains, the bundles of telegraph wires, the
noisy cities and great mills of to-day, we can remember Philip
Schuyler, Sir William Johnson, Marinus Willett, Peter Gansevoort,
and Nicholas Herkimer. There were no nobler patriots, even in
Virginia and Massachusetts, than these men of the Mohawk valley.
 CHAPTER IV
THE ERIE CANAL

If we think that the men of a hundred years ago were people with
few wants, who were willing to let others do the trading and make the
fortunes, we are quite in the wrong. They were as eager in business
as are the driving Americans of to-day. So long ago as 1683 Thomas
Dongan, a well-born Irishman, came to New York to be its governor.
In his letters to the government in London he said a great deal about
the fur trade and the danger of its going to other cities. Once he
reported that two hundred packs of beaver skins had gone down the
Susquehanna river and across to Philadelphia instead of being
brought by the Mohawk to New York, and he thought that if this traffic
continued New York would be ruined.
As time went on the rivalry grew stronger and stronger. All the
cities on the coast were bidding for the western trade. The “West”
was then the Genesee country, the plains along the Lakes, and the
rich lands of the Ohio valley. Some of the trade from the Lakes and
the Genesee went down the St. Lawrence. Heavy articles especially
were sent to Quebec, while lighter freight was taken overland down
the Mohawk. When De Witt Clinton was stirring up the legislature
and the people of New York, he told them he was very sorry to learn
that merchandise from Montreal was sold in the state for less than
New York prices. This was because there was transportation by
water from Montreal, and the St. Lawrence merchants could afford to
undersell those of New York.
Many people thought that the wheat and flour and other products
of western New York would all go down the Susquehanna to
Baltimore and Philadelphia. Rough boats known as “arks” were built
and floated down the river in the high water caused by the melting of
the snows in the Allegheny highlands. From two to five hundred
barrels of flour were carried in one of these craft. As the boats could
not be sailed up the river, they were taken to pieces at the end of the
voyage and sold for lumber. We have already seen that Colonel
Rochester followed this valley in migrating to the Genesee river, and
one writer calls attention to the fact that in seven days several elderly
people had come quite comfortably by this route from Baltimore to
Bath in the southwestern part of New York. One could now travel
from San Francisco to New York and almost halfway across the
Atlantic ocean in that time.
Other cities also hoped to secure some of the profits of dealing
with the rapidly growing West. The tourist on his way down the
Potomac to Mount Vernon, the home of Washington, will pass by
Alexandria, a quiet old town of about fifteen thousand people.
Washington himself thought it possible that Alexandria might get a
good share of the trade from Detroit and other places on the Lakes
and on the Ohio river. All this seems strange to us, because since
the days of our great-grandfathers the traffic has been going largely
to New York. The cause of the change was the Erie canal. Yet in
1818, a few months after the canal was begun, an Albany
newspaper discussed very earnestly, as one of the chief questions of
the day, the danger that Philadelphia would take away the western
trade.
Flour, salt, and potash had been taken to New York in large
quantities, but all these products were carried as far as Schenectady
in little ten-ton boats, by way of Wood creek and the Mohawk. As the
business grew it was seen to be impossible always to drag the boats
up Wood creek with horses, and that the small canal, ten feet wide,
which had been cut around the rapids at Little Falls, could not serve
the purposes of another generation.
Hence for many years there had been talk of a canal to join the
Lakes and the Hudson, thus making navigation without a break from
the interior of the country to the Atlantic ocean at New York. The
credit for first thinking of such a canal has been claimed for several
men, but probably it was “in the air,” and many thought of it at about
the same time.
 Gouverneur Morris, one of the famous New York statesmen of
the day, proposed that lake Erie should be “tapped” and its waters
led to the Hudson. The surface of this lake is five hundred and
seventy-three feet above tide water at Albany. It was Morris’s idea to
dig a channel, with a gently sloping bottom, which should send the
water east in a stream deep enough to float a boat. The water thus
turned from its course would go to Albany instead of flowing through
the Niagara and the St. Lawrence. There were, however, difficulties
about the plan which Morris did not understand, and it was never
carried out.

Fig. 13. De Witt Clinton

The great water way is often known as “Clinton’s Ditch.” This
name was doubtless given in ridicule by those who did not think it
could be built. There were many who laughed at the surveyors when
they saw them looking about, using their levels, and driving their
stakes in the woods and swamps. It was even said that to dig such a
canal was impossible, that it would cost too much money, that it
would take too much time, and that the canal itself could never be
made to hold water.
But Clinton and his supporters believed in it, and worked hard to
make it a success. They said that the cost of carrying a ton of
produce in wagons a distance of one hundred miles was about thirty-
two dollars. The experience of others had proved that in canals a ton
could be carried one mile for one cent, or a hundred miles for one
dollar. There is a great difference between one dollar and thirty-two
dollars, especially if the difference is added to the cost of the wheat
from which our bread is made, or of the lumber used in building our
houses. Clinton himself thought that it might take ten or fifteen years
to make the canal, but, as we shall see, it was finished in less time
than he supposed.
Clinton declared very truly that New York was especially
fortunate, for the surface made it an easy task to dig the ditch. There
was no high or rough ground to be crossed, there was plenty of
water to keep the canal full, and it would run through a fertile and
rich country. As Clinton was governor of New York during much of
the period in which the canal was made, his name is imperishably
connected with the great enterprise. He was once candidate for the
office of President of the United States, but perhaps even that office,
if he had been elected, would not have given him so much honor as
did the building of this great public work.
Canals were not new in Clinton’s time. Long before the Christian
era began men had dug them to carry water for various uses, such
as irrigation and turning machinery. Often, as for hundreds of years
in the fen country of England, canals have been used to drain wet or
flooded lands and for moving boats. Even beavers have been known
to dig ditches, which fill with water, that they may float the wood
which they cut to the place where they build their dams and their
homes.
If a region is perfectly level, only a ditch and water are needed.
But lands are not often level for more than short distances; hence a
canal consists commonly of a series of levels at different heights. Of
course the boats must be passed from one level to another by some
means. If they are small, they can be dragged up or down between
two levels; but this method will not serve for large boats carrying
many tons of coal, lumber, salt, or bricks, hence locks are generally
used. A lock is a short section of a canal, long enough for the boats
used, and having walls rising from the bottom of the lower level to
the top of the upper one. There are big gates at each end. If a boat is
to ascend, it runs into the lock on the lower level and the lower gates
are closed. A small gate in the large upper gate is then opened and
the water runs in from above, slowly raising the water in the lock and
with it the boat. When the water in the lock is even with the water in
the upper level, the big upper gates are swung open and the boat
goes on its way. In a similar manner boats go down from higher to
lower sections of the canal. Locks have been used in Italy and in
Holland for more than four hundred years.
On April 15, 1817, the legislature passed the law for the
construction of the long ditch, and the first spade was set into the
earth by Judge John Richardson at Rome, New York, on July 4 of
the same year. This was forty-one years after the Declaration of
Independence, and it is plain that the country had grown much in
wealth and numbers when a single state could start out to build a
water way three hundred miles long. After the first spadeful of soil
had been lifted, the citizens and the laborers eagerly seized the
shovels, and thus everybody had a small share in beginning the
great work. Guns were fired and there was much rejoicing.
 Fig. 14. Erie Canal, looking East from Genesee Street
Bridge, Utica
The men who took the contracts for digging short sections of the
canal were mainly farmers who had gained good properties and who
were living along the line. In those days, if any one had visited the
men at work, he would not have seen crowds of foreign laborers
living in huts, but men born and reared in the country round about. It
was little more than twenty years since the Genesee road had been
built through central New York, and there was still much forest. The
trees grew rank and strong, and it was no light task to cut through
the tangled network of roots that lay below the surface. First the
trees were cut down, making a lane sixty feet wide, and in this the
canal was dug to a width of forty feet. Powerful machines that could
draw out stumps and pull over the largest trees were brought from
Europe. The wheels of the stump machine were sixteen feet across.
A plow with a sharp blade was also made, to cut down through the
heavy carpet of fibers and small roots.
 Swiftly one piece after another of the canal was finished and the
water let in. The trench was found to hold water, and boats were
soon busy hauling produce from town to town. In 1825 it was
finished from Black Rock, or Buffalo, to Waterford, above Troy. The
work had taken eight years and had cost a little less than eight
million dollars. De Witt Clinton was right and the croakers were
wrong. Perhaps it was hard at that time to find any one who did not
think that he had always wanted a canal.
There were, it is true, a few disappointed ones at Schenectady.
There the wagons from Albany had always stopped, and there the
boating up the Mohawk had begun. As all the loads had to be shifted
between the river and the land journeys, there had been work for
many men. Thus the place had grown up, and now that boats were
to run through without change, some people naturally thought that
the town would die out, or would at least lose much of its business.
These few discontented folk, however, were hardly to be counted,
among the thousands who exulted over the completed canal.

Fig. 15. Along the Canal in Syracuse

Copyrighted, 1899, by A. P. Yates, Syracuse, N.Y.
A great celebration was arranged, and the rejoicings of the
beginning were redoubled in the festivities at the end. Boats were
made ready at Buffalo to take Governor Clinton and the other guests
to New York. When the first boat entered the canal from lake Erie a
cannon was fired. Cannon had been set within hearing distance all
the way to the sea along the line of the canal. This way of sending
news was the nearest approach to the telegraph at that time. Soon
the tidings of the great event came booming down among the cliffs of
1
the Hudson and reached New York.

1
The time allowed for the signaling from
Buffalo to Sandy Hook was one hour and twenty
minutes. This programme was substantially
carried out. From Albany to Sandy Hook only
twenty minutes were required.

Two kegs of lake Erie water were put on one of the boats at
Buffalo, and we shall see what was done with them. There were also
two barrels of fine apples which had been raised in an orchard at
Niagara Falls. These were not to be eaten on the way, one barrel
being for the Town Council of Troy, and the other for the city fathers
of New York. Many people on both sides of the ocean are still eating
fine apples from the trees of the Genesee country.
One boat in the little fleet was called Noah’s Ark, and on board
were two eagles, a bear, some fawns, fishes, and birds, besides two
Indian boys. These were sent to New York as “products of the West.”
At every town there was a celebration, and great was the excitement
in such cities as Rochester, Syracuse, Utica, and Albany. There were
salutes and feasts and speeches and prayers, and the gratitude and
joy of the people fairly ran over. The greatest celebration of all was in
New York, where everybody turned out to do honor to the occasion.
The fine ladies boarded a special boat, and the “aquatic procession”
went down through the bay to Sandy Hook. It was arranged that a
messenger of Neptune, the sea god, should meet the fleet, inquire
their errand, and lead them to his master’s realm. Here Governor
Clinton turned out the lake Erie water from the two kegs into the sea
as a symbol of the joining of the lakes and the ocean. Then all the
people went back to the city and had speeches and parades, feasts
and fireworks, while the city-hall bell was rung for several hours. The
illumination was said to be a fine one, but perhaps their lamps and
candles would now look dim.
After the canal was finished the carrying business was quite
made over. Little was heard then about sending western New York
fruit and grain to Philadelphia or Montreal or Alexandria. Freighting
was so cheap that a man who had been selling his wheat for thirty
cents a bushel now received a dollar for it. In the war with England,
only a few years before, it had cost more to carry a cannon from
Albany to Oswego than it had cost to make it. The journey had now
become an easy and simple matter. Two farmers built a boat of their
own, loaded it with the produce of their farms, and took it down
Seneca lake and all the way to New York. They were let out of the
woods into the wide world.

Fig. 16. Traveling by Packet on the Erie Canal

The canal was not entirely given up to the carrying of freight.
People thought that it was a fine experience to travel in the
passenger boats, which were called “packets.” These were
considered as remarkable as are the limited express trains of to-day.
The speed allowed by law was five miles an hour. To go faster would
drive the water against the banks and injure them. The fare was five
cents a mile including berth and table. It was said that a man could
travel from New York to Buffalo with “the utmost comfort” and without
fatigue. The journey cost eighteen dollars, and only took six days!
We, of course, cannot help thinking of the Empire State Express,
which leaves New York at 8.30 a.m. and arrives in Buffalo at 4.50
p.m.

Fig. 17. Erie Canal and Solvay Works, Syracuse

If the journey of those days seems long to us, we must
remember that to most of the travelers the scenery was fresh and
interesting, for it was a visit to a new land. The rocky highlands, the
blue Catskills, the winding Mohawk, and the towns and farms of the
interior were perhaps as full of interest as the morning paper is on
the trains of to-day. From Utica to Syracuse, more than fifty miles, is
one great level; but on nearing Rochester the canal follows an
embankment across a valley, and the passengers in those days
looked wonderingly down on the tops of trees. At Lockport they
heard the clatter as they slowly rose by a long row of locks to the top
of the cliffs, and at Buffalo they looked out on a sea of fresh water. At
Utica, Rome, Rochester, and other places, after a few years, side
canals came in from north and south, from Binghamton and from the
upper valley of the Genesee; and up in the hills great reservoirs were
built, with shallow canals known as “feeders” leading down to the
main trench. These were built to make sure that there should be
water enough for dry seasons; for locks will leak, and whenever a
boat locks down a lockful of water goes on toward the sea.
Now all was stir and growth. Buffalo had started on its way to
become a great city. Rochester ground more wheat and Syracuse
made more salt. There was no doubt that New York would soon be
known as the metropolis of the western world, and “Clinton’s Ditch”
became the most famous of American canals.
 CHAPTER V
THE NEW YORK CENTRAL RAILWAY

The Erie canal had not long been finished when a new way of
carrying men and merchandise came into use in New York. In the
next year after the great celebration the legislature granted a charter
to build a railroad from Albany to Schenectady. It is sometimes said
that this was the first time in America that cars were drawn by means
of steam. This is not true, but New York was not far behind some
other states, and the De Witt Clinton train, of which a picture is
shown in this chapter, looks as if it must have been one of the very
earliest ones. This train made its trial trip in 1831, which was
seventeen years after George Stephenson had built his first
locomotive in England.
A railroad had been opened from Baltimore, a few miles to the
west, the year before, and about the same time another was built in
South Carolina. Two years earlier, in 1829, the Delaware and
Hudson Canal Company brought from England three locomotives,
one of them built by Stephenson, to draw coal to their canal from
their mines at Honesdale, Pennsylvania. In 1826 a railroad four miles
long was built at Quincy, Massachusetts, to carry granite from the
quarries to the sea. It was called a tramway, and horses were used
instead of steam. If we go to England, we shall find that tramways
have been used there for more than a hundred years. Thus it is not
easy to say when the first railroad was built, and all writers do not tell
the same story about it, but it is certain that steam cars were first
used and long roads with iron tracks were first built a little less than a
hundred years ago.