Michael Hammond. The Phonology of English: A Prosodic Optimality-Theoretic
Approach. Oxford:Oxford University Press. 1999. 368pp.
Reviewed by David Deterding
Synopsis
This book presents a detailed analysis of the syllable structure and stress
placement of English words, using Optimality Theory (OT) as the theoretical
framework for the analysis. Because of the complexity of these issues, the
consideration of stress placement concentrates almost entirely on
monomorphemic words, so only seven pages (pp.322-329) are devoted to the
influences of suffixes on stress placement, and issues of what happens to
words in context are not covered.
Chapter 1 introduces the way that OT provides a mechanism for considering
the interaction of various phonological constraints, using aspiration and
vowel nasalisation as processes to illustrate the use of OT tableaux.
Chapters 2 to 4 then consider the structure of the English syllable, with a
detailed analysis of phonotactic constraints and also the use of moras to
express constraints on syllable structure. Chapters 5 to 8 discuss the
theory of the foot and the complex interaction of various constraints in
determining the placement of stress in monomorphemic English words.
Finally, Chapter 9 presents a brief overview, and also a consideration of
further areas that deserve to be researched.
Critical Evaluation
(Note: Hammond uses [s-hacek, z-hacek] for the coronal-dorsal fricatives of
English. For typographical reasons, here I will instead use the pseudo-IPA
symbols [S, Z] for these sounds.)
Hammond's book is written very clearly, with issues concerning the OT
representation of syllable structure and stress placement in monomorphemic
English words examined meticulously and exhaustively. The book thus not
only presents a comprehensive analysis of these complex issues, but it also
provides a welcome overview of OT, and thereby demonstrates the power of
this theoretical framework. This book is certain to become an important
work on English phonology, and I think everyone should read it.
In the following paragraphs, I will raise some issues that I feel merit
further discussion. In many ways, my discussion of these issues may
actually be regarded as a tribute to the clarity of the book: it is always
highly informative and thoroughly thought-provoking, and the fact that it
encourages one to delve further into many of the issues and maybe question
some of the conclusions is a result of the overall lucidity of the
presentation.
Although most issues are clearly and carefully presented, there are a few
which are introduced a little abruptly. For example, it is a bit alarming
to be told (p.172) that the Nonfinality Constraint "prevents the final
syllable from being footed", as any speaker of English will immediately be
able to think of plenty of examples where the final syllable of a word is
not just footed, but stressed as well. In this and other similar cases,
such abrupt presentation of an idea that initially seems counter-intuitive
is mitigated by the subsequent meticulous use of a large number of OT
tableaux to show how the various constraints interact to allow the correct
derivation of the metrical stress structure of words. Indeed, after this
detailed presentation, not only does the nature and ranking of the
constraints become apparent, but the strength of the OT framework emerges
extraordinarily clearly.
Sometimes, however, the subsequent explanation of an issue really is delayed
too long. We are told (p.250) that there are virtually no clear cases of
disyllabic monomorphemic words ending in a velar nasal. On reading this,
words like 'pudding' and 'herring' immediately come to mind, and it is only
23 pages later (p.273) that we are told that such words are being treated as
if they have an -ing suffix. Not only is this delay unfortunate, I also
don't find the treatment totally convincing: while it would seem perfectly
reasonable to argue that the frequency of the -ing suffix has had a
substantial effect on the syllable structure of English words, including
monomorphemic words, I have residual doubts about the suggestion that
'pudding' and 'herring' somehow consist of two morphemes, or indeed that
'honest' and 'earnest' might also be treated as polymorphemic because they
have an -est suffix (p.252).
One final comment about the abrupt introduction of new concepts: this book
assumes quite an extensive background knowledge of phonetics and some
phonology. While it might be reasonable to assume that a reader is already
familiar with aspiration and anticipatory vowel nasality, the introduction
of the 'sonority hierarchy' with no explanation (p.85: "I propose that
margins are subject to the sonority hierarchy") might prove something of a
barrier to those not familiar with such concepts -- surely a brief
definition of sonority would be useful.
This book covers some things in impressive detail, while other matters are
perhaps skipped over rather lightly. For example, it is simply assumed that
feet in English are trochees (one stressed syllable followed by one
unstressed syllable), and no consideration is given to the possibility of
dactyls (one stressed syllable followed by two unstressed syllables) even
though this term is introduced on p.150. As a result, when all the
potential ways of parsing a 3-syllable word such as 'Christopher' are
listed, a dactylic analysis is not considered, and the final analysis that
emerges consists of a trochee followed by an unparsed final syllable. While
there is (no doubt) plenty of evidence to support this analysis, as it is
presented in this book it is not totally convincing. Indeed, while the OT
tableaux illustrate admirably how the constraints can be ranked to achieve
the correct parsing, the nagging suspicion remains that some of the
constraints are ad hoc: not only is there the failure to consider analysis
in terms of a dactyl, and also the abruptness of the presentation of the
Nonfinality Constraint discussed above, but I additionally feel the
suggestion that many verbs and nouns have a catalectic (silent) suffix
(p.278) needs further justification to supplement the demonstration that the
existence of such a suffix can provide the right answers for the stress of
English words.
Another issue that I feel could benefit from further attention concerns the
status of the coronal-dorsal fricative [Z]. As evidence for the existence
of the syllable, it is claimed that medial phonotactic constraints emerge
from a combination of syllable-final and syllable-initial constraints, but
the status of [Z] in English would seem to represent something of a
question-mark regarding this claim. Clearly there is no problem with this
sound in medial position ('measure', 'leisure'), but in initial position it
only occurs in a few borrowed words ('genre', 'gigolo') and in final
position in a few more borrowed words ('mirage', 'beige'). Hammond deals
with this by simply assuming that it belongs in syllable-final position, and
he lists 'liege', 'beige', 'rouge', 'loge', and 'garage' as words ending in
[Z] (p.112). Well, 'beige', 'rouge', and 'loge' all seem to me quite
clearly to be borrowed, while 'liege' and 'garage' both have alternatives
with [dZ], which strongly suggests the instability of [Z] in final position
(while there seems to be no such pressure on the medial [Z] in 'measure' or
'leisure'). I do not wish to suggest that Hammond is necessarily wrong in
his analysis, as there are certainly good reasons to regard [Z] as a
syllable-final consonant. However, I would have preferred that this problem
with the status of [Z] had been mentioned, even if it were simply to
acknowledge its existence.
My final comment concerns a suggested omission. In establishing the
phonotactic relationships of the English syllable, a vast and impressive
array of examples are considered, including such rare words as 'knish' and
'tmesis'. The danger is that this lays the analysis open to the 'what
about' objection. My suggestion is: what about 'Sri Lanka'? This is
relevant, because [S] and [s] are claimed to be in complementary
distribution in initial clusters, with [S] occurring only before [r], while
[s] occurs before all other consonants (p.55). While exceptions of Yiddish
origin do get mentioned, 'Sri Lanka' does not, and it would certainly appear
to be a possible exception (depending on how the first sound is actually
pronounced). Now you might object that this is a name, so it doesn't count.
But 'Minsk' gets two mentions (pp.65 & 95), and some of the names considered
can be really quite obscure: 'Bhutatathata', 'Haleakala', 'Kichisaburo',
'Anaxagorean', 'Antananarivo' .... all on p.300. So why not 'Sri Lanka'?
Summary
I have discussed a number of issues, particularly concerning the brevity
with which some issues are considered, and this might seem to suggest that
the book has severe flaws. This is not true at all: in fact it presents a
complex phonological framework (OT) delightfully clearly, and I thoroughly
enjoyed reading it, even if I might have preferred one or two issues to be
covered in more detail.
One must accept, however, that not all issues can be covered in depth, and
the impressive clarity with which Hammond has presented the phonology of
English words within the OT framework is highly commendable.
I highly recommend this book as exceptionally valuable for all students of
phonology.
[About the reviewer: David Deterding teaches phonetics, syntax, and
translation at the National Institute of Education, NTU, Singapore. His
webpage is: http://www.soa.ntu.edu.sg:8080/ell/DavidD/Personal/david.htm]
|
