LINGUIST List 12.3154

Thu Dec 20 2001

Review: Handbook of Morphology, 2nd rev.

Editor for this issue: Terence Langendoen <terrylinguistlist.org>


What follows is another discussion note contributed to our Book Discussion Forum. We expect these discussions to be informal and interactive; and the author of the book discussed is cordially invited to join in. If you are interested in leading a book discussion, look for books announced on LINGUIST as "available for discussion." (This means that the publisher has sent us a review copy.) Then contact Simin Karimi at siminlinguistlist.org or Terry Langendoen at terrylinguistlist.org. Subscribe to Blackwell's LL+ at http://www.linguistlistplus.com/ and donate 20% of your subscription to LINGUIST! You get 30% off on Blackwells books, and free shipping and postage!

Directory

  1. Phoevos Panagiotidis, Review of Spencer & Zwicky, eds., Handbook of Morphology

Message 1: Review of Spencer & Zwicky, eds., Handbook of Morphology

Date: Thu, 20 Dec 2001 15:34:35 -0000
From: Phoevos Panagiotidis <phoevosntlworld.com>
Subject: Review of Spencer & Zwicky, eds., Handbook of Morphology

Spencer, Andrew, and Arnold M. Zwicky, ed. (2001) The
Handbook of Morphology. Blackwell Publishers, paperback
ISBN 0-631-22694-X, xvi+815pp, $49.95. Blackwell Handbooks
in Linguistics (hardback ISBN 0-631-18544-5, 1998).

Phoevos Panagiotidis, University of Cyprus

[For another review of this book see
http://linguistlist.org/issues/12/12-3096.html --Eds.]

OVERVIEW
The Handbook of Morphology (henceforth: 'Handbook') is a
collection of contributions on morphology. The Handbook
opens with lists of contents, contributors and
abbreviations and an introduction by the editors reviewing
the thematic areas the book covers, rather than linearly
presenting the contributions. The convention of marking the
titles of the Handbook's contributions in small capitals is
introduced here, which makes cross-referencing easy
throughout. The editors' introduction concludes with a list
of morphological phenomena and what languages they are
illustrated by in the Handbook.

The Handbook's contributions / chapters are divided into
five large parts. Part I is dubbed 'The Phenomena' and
comprises the following chapters: 'Inflection' by G. Stump,
'Derivation' by R. Beard, 'Compounding' by N. Fabb,
'Incorporation' by D. Gerdts, 'Clitics' by A. Halpern,
'Morphophonological Operations' by A. Spencer and
'Phonological Constraints on Morphological Rules' by A.
Carstairs-McCarthy. It is evident what each chapter is
about.

Part II is titled 'Morphology and Grammar' and includes
'Morphology and Syntax' by H. Borer, 'Morphology and
Agreement' by G. Corbett, 'Morphology and Argument
Structure' by L. Sadler and A. Spencer, 'Morphology and the
Lexicon: Lexicalization and Productivity' by M. Aronoff and
F. Anshen, 'Morphology and Lexical Semantics' by B. Levin
and M. Rappaport-Hovav and 'Morphology and Pragmatics' by
F. Kiefer. In this part various aspects of the relationship
between morphology and other components / (sub)systems of
language are examined in detail.

Part III is called 'Theoretical Issues' and it contains the
following contributions: 'Prosodic Morphology' by J.
McCarthy and A. Prince, 'Word Syntax' by J. Toman,
'Paradigmatic Structure: Inflectional Paradigms and
Morphological Classes' by Carstairs-McCarthy and R.
Sproat's 'Morphology as Component or Module: Mapping
Principle Approaches'. As is obvious, the chapters in this
part each discuss a number of theoretical issues and
approaches to morphological research.

Part IV, 'Morphology in a Wider Setting', is a collection
of contributions on the relationship or interaction of
(theoretical) morphology with other disciplines in the
wider context of Linguistics and Cognitive Science. B.
Joseph's 'Diachronic Morphology' is about 'morphology in
historical linguistics', E. Clark discusses 'Morphology in
Language Acquisition', W. Badecker and A. Caramazza review
what acquired language disorders may reveal about
morphology in 'Morphology and Aphasia'. Psycholinguistics
in the narrower sense (the study of language processing) is
covered by J. McQueen and A. Cutler's 'Morphology in Word
recognition' (perception) and J.-P. Stemberger's
'Morphology in Language Production with Special Reference
to Connectionism'.

Finally, Part V contains a number of 'Morphological
Sketches of Individual Languages'; these are: Archi (by A.
E. Kibrik), Celtic (by J. Fife and G. King), Chichewa (by
S. Mchombo), Chukchee (by I. Muravyova), Hua (by J.
Haiman), Malagasy (by E. Keenan and M. Polinsky), Qafar (by
R. Hayward), Slave (by K. Rice), Wari' (by D. Everett) and
Warumungu (by J. Simpson).

EVALUATION
The edition reviewed here is the paperback version (2001)
of the original edition (1998). It appears to be the case
that apart from the format and, naturally, the price, there
are no other differences between the texts of the 1998 and
the 2001 version of the Handbook. Because of this fact, the
book has been adequately and proficiently reviewed in
detail more than once, extracts from some favourable
reviews can be found on the back cover. Despite this fact,
I did choose to review the book as someone trained as a
syntactician, as well as someone with only selective
interests in morphology, as is expected by a non-
morphologist. At the end of the day, my aim will be to
examine to what extent the Handbook can offer someone who
has received, or is still receiving, his or her training as
a linguist an accessible, broad as well as in depth
overview of the field of morphology. Such an expectation, I
understand, is straightforwardly entertained by someone
approaching a handbook on a topic, as opposed to a textbook
or a collection of specialised or technical papers.

The above made clear, I will proceed into evaluating first
the individual contributions; when relevant I will work in
a 'comparative' fashion, as some contributions of the
Handbook are more than superficially related to each other.
Finally, I will say a couple of words about the book in
general (organisation, choice of topics, editorial work and
so on).

The Handbook opens with Stump's chapter on inflection. The
chapter can be roughly divided into three sections: the
first debates the status of inflection (as contrasted with
derivation); the second reviews properties and functions of
inflection cross-linguistically; the third reviews
theoretical approaches to inflection. The discussion is
well organised and informative throughout, clear but never
oversimplified. In pp.14-19 Stump deconstructs the
'traditional' (i.e. 'theory-neutral') criteria for
distinguishing between inflection and derivation and shows
them to be inadequate, he nevertheless concludes by
stressing the necessity of coherent formal analyses in
order to capture and establish this distinction in a
fruitful and empirically useful fashion.

A chapter on derivation naturally follows, this one by
Beard. The chapter comprehensively presents not only the
different varieties of derivational morphology attested,
but also the main theoretical models to account for them.
Although the contribution remains more or less
theoretically not committed, it upholds the necessity of a
distinction between derivation and inflection more
confidently than Stump's, although both chapters stress the
contradictory predictions of the usual diagnostics
distinguishing between the two. Interestingly, Beard draws
the line on p.56-7 between derivation, a language-internal
process, and extragrammatical lexical stock expansion, like
borrowing ('troika'), acronyms ('UN') or conscious
analogical coining of new bases ('chicken-burger' from
'hamburg-er'). I found Beard's stressing the difference
between derivation as a grammatical process or operation
and conscious coinings most welcome and the reason for this
can be succinctly expressed by the truism that unless the
field of enquiry of a (sub-)discipline, such as morphology,
is carefully delineated, no useful or interesting
generalisations or theory building can be made.

Fabb's chapter on compounding makes an outstanding
contribution to the volume and this is not exclusively down
to the fact that it has a narrower scope than the previous
two chapters, which face the much more daunting task of
discussing very broad issues like 'inflection' and
'derivation'. The discussion of different types of
compounding and the variety of relationships between their
constituent parts is constructive and comprehensive, as
compounding is discussed in relation with 'compound-like'
constructions and issues in the phonology of compounding
are also presented. My only objection is that a paragraph,
or an endnote, dedicated to 'psycholinguistic' research on
compounds would not be completely out of place and would
probably further illustrate the affinities between
derivation and compounding. An example is Gordon's (1986)
experiment, which clearly demonstrated inflection not being
tolerated inside English compounds (*rats eater) by 3 to 5-
year-olds, although both 'uninflected' nouns in singular
(rat eater, mouse eater) and compounds with irregular
plurals (mice eater) were fine.

Gerdts offers a discussion of incorporation, where not only
the syntactic restrictions on the phenomenon are presented,
but also similar phenomena like Tongan noun stripping,
which does not straightforwardly involve a morphological
operation next to the object appearing to the left of the
verb. When discussing the status of the incorporated noun
stem on pp.85-86, she notes that the more 'salient' a noun
is, the least likely it is to incorporate, while Southern
Tiwa proper names, for instance never incorporate. It is
also stated that incorporated nouns are "devoid of
discourse focus". Given also the very interesting
observation on the top of p.85 that only noun stems
incorporate in Chukchee as well as that clauses with
incorporation have a generic reading, we would be much
better off arguing that only indefinite nouns, nouns not
headed by determiners and/or not marked for case, can be
incorporated. This would naturally account for the
genericity, and recourse to notoriously elusive notions
such as 'salience' and 'discourse focus' would be avoided -
- given that we appear to have a more concrete alternative
here.

Halpern's contribution on clitics is succinct but
interesting, extremely informative and illustrated well
with examples. Halpern knows better than focus exclusively
on pronominal clitics, but he covers their positioning
adequately and even portrays the phenomenon of clitic
doubling (as in the Spanish 'Lo visto a el'). He discusses
the virtues and shortcomings of a Klavans (1980) style
account of the positioning options of clitics with
admirable insight and clarity. Apart from an unfortunate
typo on p.104, where in examples (11) and (12) French 3rd
person 'voit' has become 'vois', this chapter is an
excellent presentation of clitics and an ideal starting
point for anyone making a first encounter with their
strange and exciting properties.

Spencer's chapter is more theoretically oriented. This is a
most welcome fact, not only because of the soundness of his
arguments, whether one agrees with the conclusions or not,
but also because Spencer demonstrates his commitment to a
view of morphology being somehow mentally represented,
rather than an abstract paradigmatic curiosity fit for
taxonomic approaches only. This becomes even more important
given the typological thoroughness of his argumentation.
The main argument in this chapter is against morphemes as
minimal meaningful forms, and the concomitant Item-and-
Arrangement view of morphology, and for a rule-based one in
the spirit of Item-and-Process being descriptively
superior; I would nevertheless wish to focus on his
arguments on pp.126-7 about how we can decide which view is
explanatorily superior. Spencer starts off from the common
view that Item-and-Arrangement morphological models are
superior because they are more restrictive, whereas rule
systems tend to be freer. He further goes on to make the
very interesting observation that, unlike what is the case
with unbounded and underdetermined by the primary data
syntax, "a characteristic of morphological systems is that
they are subject to idiosyncratic restrictions, indicating
that storage, rather than generation, is the key device".
In other words, a language learner does not know the
morphological system of a language before being exposed to
all the relevant forms (whereas s/he does not have to be
exposed to parasitic gap constructions in order to produce
/ understand them). Although some rule application does
seem to be involved in morphology, especially inflection,
Spencer's claims would probably make the right predictions
for irregular inflectional morphology (which clearly has to
be learned) and most of derivation. Especially in
derivation, Spencer's view of morphology as essentially
non-generative would entail that productivity of certain
derivational affixes is generalisation rather than
generation (see also Aronoff and Anshen, this volume, and
Pinker, 1999). Spencer's being a strong and clear claim, it
remains to be empirically examined and it has to.

Part I concludes with a short discussion of phonological
constraints on morphology by Carstairs-McCarthy, revolving
around a number of relevant examples. I found the chapter
quite informative, although in its purpose it partially
overlaps with Spencer, Aronoff and Stump, all in this
volume.

Borer offers a detailed, wonderfully articulated and
comprehensive presentation of the models of the syntax-
morphology interaction in the market and she reaches the
conclusion that we cannot dispense with morphology, as
often attempted (Baker, 1988 is arguably an example of this
trend, at least an oft-cited one). The 'mixed' model she
offers lets syntax do the syntactically relevant word-
building only, in this sense the chapter opens an
interesting dialogue with Beard, Toman and Sproat (all this
volume).

Corbett's discussion of agreement is systematically
presented and a number of exciting facts about it are
summarised, moreover, the text is richly referenced for the
interested reader who would be interested in more details.
Nevertheless, the contribution only works at the level of a
review article and it tends to dodge explanation and
hypothesis formation, restricting itself to typologically
reviewing the phenomenon of agreement, although it is
refreshingly informed from Phrase Structure grammars
(p.192). Symptomatic of this tendency is the section on
pp.198-201, where we are effectively given a long list of
what can agree with what, pretty much everything reasonably
close to each other, in which environments. This is a
disappointment, especially compared with the theory-neutral
but explanation-driven discussion of gender in Corbett
(1991), the most enlightening work on the topic. Finally,
it is not straightforward at all that antecedent-anaphor
covariance is agreement proper (cf. p.191), whereas
examples of 'pragmatic agreement' (p.192), such as 'the
government are trying to send asylum seekers back' as
opposed to '*I is parked' for 'My car is parked there',
have been most recently given a syntactic account. (Den
Dikken, 2001).

'Morphology and Argument Structure' by Sadler and Spencer
is looking at morphological valency variation (i.e.
operations that decrease a predicate's arguments, like
passivisation, or operations that increase them, such as
causativisation) and is closely related to Levin and
Rappaport-Hovav (this volume). Although this a very
interesting approach to the topic, it is not always reader-
friendly and the style is occasionally too dense.

Aronoff and Anshen's chapter is delightful. Not only is it
one of the few examples of work in linguistics (summarising
of course previous work by them) where corpora are put to
very good use, that is they are used to explain facts about
human cognition (the interaction between stored words in
the lexicon and morphological operations here), but it also
offers very solid and crystal clear argumentation. Their
treatment of frequency of form and how it relates to
productivity is particularly instructive and their
correlation of productivity to the communicative use of a
derived term (jargon terms tend to use less productive
affixes: p.246) also appears quite plausible.

Levin and Rappaport-Hovav look at verbs with similar
lexical conceptual structures (e.g. verbs like 'walk', as
in 'I walked on ice' and 'I walked the distance alone',
expressing an activity and an accomplishment respectively)
and show they tend to either have an identical
morphological exponence (as English 'walk' above) or differ
minimally from each other, as in Russian and Hebrew; at the
same time it is never the case that, say, accomplishments
share any morphology in any language whatsoever by virtue
of their similar semantic template. The morphological
treatment of such matters is influential and has produced
significant results. On the other hand the picture appears
to be somehow more complex. Hence, in English it is
systematically the case that activities can be made into
accomplishments, as illustrated with 'walk' above, and as
such the alteration could also be captured syntactically,
and it has indeed (see Arad, 1996; Baker, 1997 and
references therein). At the same time, in languages like
Modern Greek the possibility of 'turning' an activity verb
into an accomplishment one heavily depends on the choice of
preposition of the accomplishment verb's complement (see
also Klipple, 1997 for related phenomena), which suggests
again a syntactic account can be relevant in Modern Greek,
as well as English, whereas a morphological one could be
suited for Russian and Hebrew.

'Morphology and Pragmatics' discusses a range of morphology
motivated by pragmatics, 'morphopragmatics' or
'extragrammatical morphology'. The impression the reader
gets is nevertheless that such a thing does not exist,
either because the presented phenomena involve
pragmatically relevant information encoded in the grammar,
like diminutives, augmentatives or honorifics, as well as
the not mentioned focus and topic particles, or because
they are matters of word choice and hence unrelated to
grammar, like when augmentatives are used. An example of
the latter is, again, Greek, where from all the 'national'
nouns (Spaniard, Pole etc.) only two can take augmentatives
and this does not appear to be a matter of morphology.

Part III opens with McCarthy and Prince's account of
prosodic morphology in infixation and reduplication. They
first employ 'hard' and unranked constraints and then
proceed to show the descriptive superiority of ranked soft
constraints, as more instances of reduplication can hence
be captured. This is a very clear and well-structured
chapter and can also serve as a cogent introduction to
Optimality Theory, making this contribution both
theoretically and descriptively appealing.

Toman's 'Word Syntax' gave me the impression of
significantly overlapping with Borer's chapter, although
the discussion of 'base-generated word structure' on
pp.313-320 is certainly not without interest.

'Paradigmatic Structure' by Carstairs-McCarthy is a review
article looking at how paradigms are structured and what
the limit of their complexity can be. On p.325 he reviews
Bybee's typological work on what categories can be
'inflectionally expressed' on the verb. After citing
percentages of which of these categories are more common,
we read that Bybee's conclusion is that 'relevant' and
'general' inflectional categories tend to be more common.
Characterising Mood (which wins with 68% in Bybee's sample
of fifty languages) as more relevant and/or more general
than Tense (5th with 48%) cannot be established on any
conceivable concrete (and falsifiable) criterion of
semantic or pragmatic/communicative nature and I am puzzled
that Carstairs-McCarthy took the trouble of reporting
Bybee's work so closely, especially given the fact that on
pp.331-4 we are presented with a very interesting
discussion of whether there is a 'default' declension class
in grammars possessing more than one.

Sproat's chapter argues for Morphology not being a real
module of grammar and presents Sadock's Autolexical Syntax.
He then goes on to analyse the bracketing paradox in these
terms (pp.342-7). The bracketing paradox is illustrated by
such comparatives as 'unhappier': although 'unhappy', like
'ecstatic' is trisyllabic, it can nevertheless form a
morphological comparative with '-er' (cf. *'ecstaticker'),
probably on the basis of 'happy'. At the same time,
'unhappier' does not mean 'not happier' (as we would
expect, given that 'happy' is responsible for the
morphological comparative), but 'more unhappy'. Sproat
points out that the inflectional '-er' is active at a
different level than derivational 'un-'

In Part IV now, Joseph's 'Diachronic Morphology' examines
in which ways morphology can change over time and what
factors trigger such change. Contrary to expectations,
there is little on grammaticalisation as such, although its
effects are presented in several instances. The exposition
is in parts confusing as the elaborate presentation of
individual facts obscures the phenomena they illustrate. I
also felt that too much space is dedicated to analogy
(although it is acknowledged on p.363 that its sources are
far from uniform) and on p.360 acronymic coinage (like UN)
is put on a par with other, non-conscious lexicalising
processes, something that does not appear to be right (see
also the discussion on Beard's and Anshen & Aronoff's
contributions above). The chapter contains a wealth of
linguistic evidence and a clearer presentation would make
it more accessible, as well as the diachronic trends they
illustrate.

'Morphology in Language Acquisition' claims that regression
in morphological acquisition is not attested (the fact that
children overgeneralise rules and produce non-existent
forms like 'goed', 'catched' and 'bringed' giving up the
previously learned 'went', 'caught' and 'brought'; p.377),
as children arguably treat irregular forms as stems with as
yet unspecified different possible meanings. On pp.378-9,
it is argued that because different children overgeneralise
at different rates, it is possible that children only apply
the rule when their memory fails to retrieve a stored
irregular form -- hence we are not dealing with an actual
'stage' in acquisition. The distinction is possibly without
substance for two reasons: first, stages in language
acquisition are not deterministic and, as is known, a lot
of individual variation exists, some children may grow out
of a particular stage earlier than others and so on;
second, if Pinker (1999) is right, in both children and
adults, morphological rules like the past /d/ one take over
when memory fails us anyway (this becomes more obvious in
languages with complex morphological systems), that's why
irregular verbs that become less frequent over time tend to
become regular. The rest of Clark's paper interestingly
shows that inflection arises in children's speech before
derivation, possibly another hint towards them being
differently represented.

A similar conclusion is drawn from the discussion of
Badecker and Caramazza regarding aphasia this time. Theirs
is a very detailed and clearly written chapter, showing how
morphology can be selectively impaired and how sometimes
aphasics can 'disitinguish' between inflection, as in 'eat-
ing an apple' and homophonous lack thereof as in 'I made a
drawing'. The research reported also suggests that words
like 'butterfly' and 'butter dish' are represented
differently, the former being unanalysable.

Findings regarding morphological processing for non-
impaired subjects are reported in McQueen and Cutler, where
evidence for a dual route mechanism for inflection is
presented. The first mechanism attempts to decompose the
form, say 'annihilate-d' and appears to be particularly
successful with rare forms, whereas the second tries to
retrieve it from memory and seems to work better with
frequent forms like 'play-ed'.

Stemberger's chapter is mainly about neural networks and
the ongoing debate of whether they can learn inflectional
morphology, the English past tense more specifically.
Having reviewed error patterns, he presents the case for
neural networks. The arguments for connectionist learning,
as elsewhere, are founded on the assumption that the human
brain (or, perhaps, its mechanism dedicated to language
learning) can use the brute computing force computers do.
This is anything but straightforward, also considering the
fact that humans possibly do not have the opportunity to be
exposed to the relevant data as many times as researchers
expose their networks in order to train them. On p.446 the
existence of a "teacher" that tells the system whether the
output is correct is defended on the basis that the
"teacher" in humans maybe "another cognitive subsystem,
possibly the comprehension system"; the question would now
be who trains the "teacher". So, even if, as Stemberger
claims, 'membled' is indeed a possible child's past tense
for 'mail', as it obeys child phonology, how does the
"teacher" / " the comprehension system" find out it is
wrong? Finally, on p.450 Clahsen et al.'s (1992) finding
that '-s' is the plural rule in German, despite the higher
frequency of '-en' is explained away by the observation
that shorter words are more frequent in German (although
the source invoked is Zipf's work in the 1930s) and that
shorter words take '-s' plurals. Again, the question is
what kind of system, other than one solely dedicated to
correlating plurals with stem frequencies, can calculate
such a correlation and implement it.

Finally, the morphological sketches of 10 languages in Part
V are a real bonus, as they provide a valuable source for
study, otherwise unavailable in most cases. For instance,
Everett's study of Wari' is only the second ever published
(p.690), whereas Kibrik's description of Archi was
translated from Russian. Rice's study of Slave Athapaskan
is also important in its theoretical findings, namely that
it is not the case that in Slave inflection is inside
derivation, among others.

To conclude the review, the editorial work is indeed superb
and I can only side with Gert Booij's phrasing that the
editors should be thanked for their work. The presentation,
including the typesetting, makes it user friendly and
readable. More importantly, not only is the concentration
of eminent contributors impressive, but also the mostly
felicitous choice of topics and the high quality of the
contributions make the Handbook an indispensable, and by
now affordable, tool for both students and researchers in
linguistics.


REFERENCES
Arad, M. (1996) A minimalist view of the syntax-lexical
semantics interface. UCL Working Papers in Linguistics 8

Baker, M. (1988) Incorporation: a theory of grammatical
function changing. University of Chicago Press

Baker, M. (1997) Thematic Roles and Syntactic Structure. In
Haegeman, L. Elements of Grammar: a handbook of generative
syntax. Kluwer

Clahsen, H., Rothweiler, M., Woest, A. and Marcus, G. F.
(1992) Regular and irregular inflection in the acquisition
of German noun plurals. Cognition 45: 225-55

Den Dikken, M. (2001) "Pluringulars", pronouns and quirky
agreement. Linguistic Review, 18: 19-41

Gordon, P. (1986) Level-ordering in lexical development.
Cognition 21: 73-93

Klavans, J. (1980) Some problems in a theory of clitics.
PhD thesis, University College London

Klipple, E. (1997) Prepositions and variation. In di
Sciullo, A.-M .'Projections and Interface Conditions'. OUP

Pinker, S. (1999) Words and Rules: the ingredients of
language. Weidenfeld & Nicolson.

ABOUT THE REVIEWER
The reviewer is Visiting Lecturer at the University of
Cyprus. He is the author of the monograph "Pronouns,
Clitics and Empty Nouns: 'pronominality' and licensing in
Syntax'", published by John Benjamins in their 'Linguistics
Today' series. His main interests include nominal phrases,
the categorial make-up of functional categories, the Syntax
of Greek, the relationship between syntax and morphology
and the acquisition and mental representation of syntactic
knowledge.
Mail to author|Respond to list|Read more issues|LINGUIST home page|Top of issue