CUNY Phonology Forum - Recently Updated Pages

CUNY Phonology Forum: Introduction

cecairns — Mon, 18 Feb 2008 05:22:10 CST

The CUNY Phonology Forum is devoted to the promotion of healthy discussion and fruitful pursuit of foundational questions in phonology. Toward that end, the Forum, in partnership with the CUNY MA/PhD Program in Linguistics, organizes conferences and makes the ideas and information developed in them available to the scholarly community. Since its inception in 2003, the Forum has organized four such conferences, the most recent of which was in January 2008.

For an introduction to the Wiki and its goals, please read the Preamble.

To engage in a discussion without editing a Wiki page, please use the Discussion Forum link at the far left. If you wish to edit or create pages, you need to be invited to join. Invitations will be made freely available to anyone. Please send an email tosyllable at cunyphonologyforum dot net and include the words "wiki invite" in the subject line; an invitation will be sent as soon as one of the organizers logs on next.

If you have any questions or comments about the Wiki itself (not concerning content per se, but more general procedural questions), please send a message to cunypfeo (Edmund O'Neill).

Song alignment question

fdell — Sat, 16 Feb 2008 17:24:23 CST

Singing and the syllable.

This question, composed by François Dell, is about how speech and song are related. In some well-known cases the syllable is the basic unit for alignment to song. Does anyone know of singing traditions in which notes of song correspond not to syllables, but to moras, or to stress feet? Please contribute to the thread about Singing and the syllable.

What can singing teach us about language? One thing singing shows is the need for the notion of syllable. Singing involves a correspondence between a text and sequence of notes (a melody); see examples in (3), (4), (10) and (18) on the handout of my presentation. [Clicking on this link will download the pdf file of Dell's handout in some browsers, and simply open it in others.] In Dell's view, this correspondence is governed by the following principle (v. (5)):

BTP (Basic Textsetting Principle): Every SYLLABLE is linked to one and only one note (but not every note needs to be linked to a syllable), and different syllables may not be linked to the same note.

In the preceding formulation of the BTP, suppose we replace ‘syllable’ by some other linguistic chunk. The chunk in question could for instance be (A) a voiced vocoid (i.e. a voiced nonconsonantal segment) or (B) a maximal sequence containing only voiced vocoids. With (A), the word “Hawaiian” must be linked to five notes: [h<ə><w><a><j><ə>n]; with (B), it must be linked to one note: [h<əwajə>n]. Such ways of singing are presumably not attested in any language.

The BTP holds for languages as diverse as English, French, Tashlhiyt Berber, Mandarin Chinese, and Tagalog. Does it hold for all languages? Let us assume that it is a fact that cases (A) and (B) are not attested among the world’s singing traditions. The chunks in cases (A) and (B), are well-defined objects, but they are not genuine linguistic units, and maybe this is the reason why cases (A) and (B) don’t exist. What about genuine linguistic units such as moras or stress feet — are notes ever aligned with these units?

Paraphonological Phenomena

cecairns — Sat, 16 Feb 2008 08:14:21 CST

The CUNY Conference on the Syllable has many presentations concerning what we might broadly call “paraphonological phenomena,” by which we mean phenomena of any kind that seem to make use of phonological representations.

Kharmalov and Cote, in their poster at the CUNY Conference on the Syllable, showed that different experimental tasks yield different syllabifications, even within subjects and using the same materials. Might this reflect the fact that we syllabify phonological representations in different ways for different tasks? Not only might the phonologists’ syllables not be the same as the phoneticians’, but in fact phoneticians might be working with different syllabifications depending on what their subjects are doing. Of course, it might also be true that syllables are different at different levels within the phonology proper.

Speech production and planning. Shattuck-Hufnagel has argued, on the basis of a broad range of data, if syllable-sized units play a role in production at all, it is in terms of stored abstract motor-control plans that are retrieved in the later stages of sound-level production planning, and subsequently adjusted to fit their prosodic contexts. She reviewed evidence from a range of sources, such as speech error patterns, rhythmic timing adjustment and cross-word-boundary consonant releases.

Temporal coordination of articulatory gestures. This was discussed by Jason Shaw in his poster, as well as in the Friday Discussion.

Singing. Dell has shown that singing is a wonderful example of a paraphonological phenomenon that not only makes reference to syllables, but can manipulate syllables for the sole purpose of creating a successful allignment of song and speech. There is a serious question whether or not languages ever use linguistic units other than the syllable for aligning song. Click here for this question and threads in response to it.

Reading. Liberman and others have shown that there is reason to believe that syllables play a role in learning to read. Gnanadesikan's presentation at the Syllable Conference explored that notion more thoroughly.

Links to web sources on related reading materials:
http://people.fas.harvard.edu/~bdsamuel/pdfs/akkad-handout.pdf
http://people.fas.harvard.edu/~bdsamuel/pdfs/akkadian.pdf

Frequency and priming effects. Joana Cholin has argued that syllables are stored in long term memory, as evidenced by frequency and priming effects.

McGurk effect. Ali et al argue that studies of the McGurk effect throw light on the internal structure of the syllable. (By the way, Wikipedia has an interesting page on the McGurk effect.)

Listening to languages other than your own. Ali et al also argue that Anglophone perceptions of Arabic also reveal syllable structure.

Word games. Numerous authors have justified various theories of syllable structure on ludlings.

Preamble / Opening Remarks

cecairns — Mon, 04 Feb 2008 06:51:45 CST

WHAT IS THE MOTIVATION FOR THE SYLLABLE IN PHONOLOGICAL THEORY?

Chuck Cairns and Eric Raimy

THIS WIKI IS AN OUTGROWTH OF THE FRIDAY DISCUSSION AT THE CUNY CONFERENCE ON THE SYLLABLE, JANUARY 17 - 19, 2008. THE PURPOSE IS TO EXPLORE THE POSSIBILITY OF A CONSENSUS AMONG PHONOLOGISTS ABOUT THE GENERALIZATIONS AND PHENOMENA THAT ARE THOUGHT TO BE WITHIN THE DOMAIN OF SYLLABLE THEORY. IT IS NOT TO EXPLORE ALTERNATIVE VIEWS OF THE SYLLABLE NOR TO DEBATE WHETHER OR NOT THE SYLLABLE IS A VIABLE CONSTRUCT IN PHONOLOGICAL THEORY.

THIS OPERATES JUST LIKE ANY OTHER WIKI: ALL REGISTERED USERS MAY EDIT THIS DOCUMENT, AND EACH ACT OF EDITING LEAVES A THREAD THAT OTHERS MAY VIEW. YOU MAY BECOME A REGISTERED USER BY GETTING AN INVITATION TO JOIN. IF YOU DO NOT ALREADY HAVE AN INVITATION AND WISH TO RECEIVE ONE, PLEASE SEND AN EMAIL TO syllable [at] cunyphonologyforum [dot] net, WITH THE WORD “INVITATION TO WIKI” IN THE SUBJECT LINE. AS SOON AS ONE OF THE ADMINISTRATORS LOGS ON, YOU WILL BE SENT AN INVITATION (subject only to certain reasonable conditions, of course).

Introduction. Participants at the Friday Discussion at the recent CUNY Conference on the Syllable proposed that we post a document about the reasons for the syllable in phonological theory; this is to serve as a basis for discussion, which we will conduct via this “Wiki” that Edmund O’Neill has created for us. This part of the Friday Discussion was triggered by Paul Kiparsky’s challenge:

I’m wondering if we could agree on some core phenomena or generalizations that would be important to address that have been, at least on some syllabic theories, understood to be explained by syllable structure. So, anyone who proposes a theory of the syllable or … a theory of the nonsyllable, should provide an account of those things. Those might include, for example, the well established typological generalizations, … generalizations about phonotactics, about compensatory lengthening, and maybe some core facts about English, or other well known languages… let’s say Kahn’s stuff. People who introduce new ideas should take care to have a story about those. Otherwise we end up talking past each other…

This brief essay is an attempt to outline the phenomena and generalizations that most contemporary theories of the syllable have traditionally been proposed to account for. It is based in part on comments drawn from the Friday Discussion, emails that colleagues have sent us since the conference, and an extremely cursory glance at the phonological literature.

First, an important caveat is in order. Facts do not come prelabeled for the theories that should explain them. It might very well be that the facts referred to below as originally motivating syllable theory are better accounted for with entirely different theoretical proposals; as progress is made in our disciplines, perhaps we will witness a chipping away at the mound of evidence once thought of as solidly supporting the syllable so that we are left with essentially nothing.

Another caveat, most forcefully expounded by Bill Idsardi, is that the enterprise embodied in the writing of this essay is essentially conservatizing. It is important to take this essay not as a prescriptive list of explananda. Rather, this essay is more in the spirit of taking stock of syllable theory in 2008. Syllables have been proposed as the explanans associated with a number of explananda. What are these explanans and explananda, and is the association between them valid?

Stefanie Shattuck-Hufnagel, in a recent email, suggested that we divide the discussion into three parts, as paraphrased below:

1. What is the range of grammatical/linguistic phenomena we have traditionally thought of as requiring the syllable for an adequate description?

2. What other theoretical structures might be able to account for these grammatical phenomena?

3. To what extent are the grammatical notions of the syllable also able to account for behavioral phenomena that have also traditionally been described in terms of the syllable? e.g. apparently-syllable-manipulating language games, apparent syllable frequency effects, constraints on co-articulatory overlap, etc.

We will in fact give scant reference in this essay to question 2, as our purpose is to provoke colleagues to suggest and debate new theoretical entities. We will, therefore, simply list some key topics from (1) and (3); we will refer to the latter as “paraphonological phenomena.”

PLEASE BEAR IN MIND THAT THE PAGES HERE (GRAMMATICAL PHENOMENA AND PARAPHONOLOGICAL PHENOMENA) ARE LISTS OF STATEMENTS DESIGNED TO PROVOKE DISCUSSION. IT IS IN NO WAY A LIST OF DEFINITIVE CLAIMS! WE HAVE SIMPLY LISTED ONE OR TWO KEY PHENOMENA THAT AUTHORS HAVE SUGGESTED RELY CRUCIALLY ON THE SYLLABLE. CONTRIBUTERS TO THIS DISCUSSION MAY SUGGEST BETTER EXAMPLES AND/OR ANTI-SYLLABIC ANALYSES OF THESE EXAMPLES.

Grammatical Phenomena

cecairns — Mon, 04 Feb 2008 06:25:39 CST

The brief survey below refers to a few examples in the literature where the syllable has been invoked to explain linguistically significant generalizations. Can the syllable be dispensed with in these and other cases without loss of descriptive or explanatory adequacy? If not, how much internal structure is required?

Reduplication. Jason Haugen argued in his talk at the conference that the syllable is a delimiter for reduplication. Can a theory of the nonsyllable offer as good an account of his facts?

Allomorphy. Vaux (Bert Vaux 1998. The phonology of Armenian. Oxford: Clarendon Press) argues that some dialects of Armenian choose the plural allomorph depending on whether or not the base is greater than one syllable.

Phonotactics. Pat Shaw argued that the notion of the syllable is not relevant (and therefore otiose) to describing consonant clusters in Salish languages. However, Toni Borowsky (Toni Borowsky 1989 “Structure preservation and the syllable coda in English” Natural Language and Linguistic Theory 7:145-166.) has argued that the word internal phonotactics of English words is best explained by reference to a maximal CVX theory of syllable structure. Kenstowicz’s text (chapter 6) contains several examples where the syllable is invoked to explain phonotactics.

Distribution of allophones. Can Kahn’s (Daniel Kahn 1976 Syllable-based generalizations in English phonology. Cambridge Mass.: MIT Ph.D. dissertation. Distributed by the Indiana University Linguistics Club.) is classic – can this be analyzed better – or at least as well – in a theory without any reference to syllable structure? Coetzee’s talk “Allophonic cues to syllabification” is also relevant to this point.

Compensatory lengthening. Paul Kiparsky’s talk on compensatory lengthening in Finnish clearly makes reference to the syllable, as do most other references to compensatory lengthening in the phonological literature.

Assimilation. The syllable is not usually thought of as necessary to accounts of assimilation; Ranjan Sen argued that it is not necessary (and therefore otiose) in accounting for diachronic processes in Romance.

Stress. The distinction between light and heavy syllables is obviously crucial for determining word stress in many languages.

Epenthesis and deletion. Syllable based generalizations about epenthesis and deletion (especially vowels and consonants, respectively) abound in the literature, especially since Itô (Itô, Junko 1989. “A prosodic theory of epenthesis.” Natural Language and Linguistic Theory. 7.217-59.) It’s hard to see how these elegant generalizations can be captured in a theory that has no syllables at all.

A significant portion of the phonological grammar of a language. Paul Kiparsky mentioned in the Friday Discussion that Panini managed an account of Sanskrit without the syllable, as did Chomsky and Halle in SPE. It is important that any theoretical construct in phonology get beyond what Idsardi (at dinner after the Friday Discussion) referred to as “helicopter linguistics,” where a few small examples – equivalent to pedagogical problem sets – are use to justify a theoretical innovation. Paul also mentioned this problem in reference to the current state of OT.

Typological generalizations. By this is meant, for example, the well known generalizations to the effect that word initial sequences seem to follow universal implicational laws. Clements (in press) has interesting comments on this.

Where did the idea of the syllable arise in generative phonology in the first place? P aul Kilparsky describes the origins of the syllable in generative phonology in the Fall of 1966.