Language Log

« previous post | next post »

This guest post by Mark Dow is an excerpt from an interview conducted via email in May/June 2021. The complete interview appears in PN Review #263 (Jan./Feb. 2022).

 Nigel Fabb is Professor of Literary Linguistics at the University of Strathclyde in Glasgow. His books include What is Poetry: Language and Memory in the Poems of the World (Cambridge 2015); with Morris Halle, Meter in Poetry: a New Theory (Cambridge 2008); Language and literary structure: the linguistic analysis of form in verse and narrative (Cambridge 2002); and Linguistics and Literature (Blackwell 1997). In 2022 he will have two new books, Thrills, epiphany, sublime: how literature surprises us. (Anthem) and with Venla Sykäri (eds.) Rhyme and Rhyming in Verbal Art, Song and Language (Studia Fennica Folkloristica).

Mark Dow is author of Plain Talk Rising (poems) and American Gulag: Inside U.S. Immigration Prisons (California).

MD: Let's start with the 'tip-of-the-tongue' phenomenon. It interests you, I think, as an example of linguistic structure detaching itself from 'content'. It points to the existence of 'unnameable' or 'unspeakable' content. On a larger scale, this feeling of knowing something we cannot put into words is what has been labelled 'the sublime.' Can you lay out the milestones of your interest in these topics?

NF: The tip-of-the-tongue feeling is a feeling relating to knowledge, and one of a number of 'epistemic feelings' which belong on what William James called the fringe of consciousness. It is related to the 'feeling of knowing', identified by J. T. Hart in 1965, which involves a feeling that we know the answer to a question, but cannot yet express it  (but are eventually able to).  I have just completed a book for Anthem Press, Thrills, sublime, epiphany: how literature surprises us, and which is about what (following Alf Gabrielsson) I call 'strong experiences'. These include experiences of suddenly feeling that we know something very important, as in the experience of the sublime and of 'epiphany' in all its variations.  Sometimes – as in aha-moments and some epiphanies – it is possible to put that knowledge into words.  And sometimes it is not. I suggest that all of these experiences begin with surprise, when we perceive something which we cannot quite fit into what we already know; and already in that, there is the possibility that what we perceive cannot be put into words, since we only have words for what we already know. There are all sorts of reasons why we cannot put into words what we think we know, and there are psychological explanations for these.  Raffman's Language, Music and Mind explores various sources of ineffability, and Sperber's On Anthropological Knowledge shows how we can fail to understand our own beliefs, including very profound beliefs, which would make them ineffable.

Are you trying to connect the measureable with the unmeasureable, the way mathematics is the basis of music and, presumably, of music's effect on us?

At least one kind of ineffability emerges from the micro-levels, in the sense that we have perceptions which are too fine-grained for us to have words for them; Raffman calls this 'nuance ineffability'.  The relation between mathematics and music's effect on us might be because we have cognitive abilities which respond to the mathematical forms underlying music; but also because we have statistical or probabilistic knowledge (which thus has a mathematical basis). That probabilistic knowledge is involved in forming expectations which can be violated or met, which is a source of feeling.

Donald Davie, in his essay 'Syntax as Music' on Susanne Langer, says that most poets who say poetry is like music don't really know music. He writes: 'In [Langer's] view a poem is like a piece of music in that it articulates itself; and in thus establishing internal relations, establishes also relations of feelings, building up the structure, the morphology of feeling, and telling us "what it feels like to feel"'.

This might be a way of saying that the meaning of the poem arises from the combination of the parts, which when taken apart do not have the same meaning (i.e., the opposite of Brecht, 'pluck a rose and every petal is beautiful'). This is true in principle for any utterance, which can always be more than the combined meaning of the parts; this is because language is partial evidence for meaning rather than fully encoding meaning. The same must be true for the non-meaning effects such as 'feelings'.  But incidentally that means that even the text as a whole is not enough, but must always be contextualized to produce meaning.

 Can you give us a lay explanation of 'generative metrics', of which Morris Halle is considered the founder, and of the the gist of your book with him? 

A generative theory is a set of instructions for building something.  A generative metrical theory is a set of instructions for building a scansion, which is a particular kind of complex structure, and that scansion is fitted to the line of poetry in a way which is also governed by instructions or rules.  Each distinct metre has its own set of instructions. Simple version: the instructions for iambic pentameter end up by building a scansion which has ten metrical positions, alternating weak (odd numbered) and strong (even numbered).  The line of poetry is fitted to the scansion, according to specific rules: the basic rules for this metre are that each syllable must fit a metrical position, which means that there are ten syllables in the line, and syllables carrying fixed stress must not match to weak positions (so for example, position 5 is weak and we would not expect a stressed syllable in this position in this metre).  If the scansion can be built and fits the line, then the line is metrical, and if not it is unmetrical (both relative to iambic pentameter, in this case).  There are further complications, for example sometimes two adjacent syllables can fit into one metrical position as in synaloepha, and sometimes there is a final extra unmatched syllable, and the rhythmic constraint is also subject to some variation.  The purpose of a generative metrics is to understand exactly what in our ordinary linguistic capabilities gets adapted to produce the modified kinds of language found in poetry, so it is part of the general linguistic project. It isn't about interpretation, and in that it is very different from a literary account of metre.

There are some big differences between this and standard foot-based approaches to metre, of the kind more commonly found in literary analysis, in which for example the iambic pentameter line is made of a sequence of five iambic feet, and feet can be substituted for other feet.  In generative metres, the line is always taken as a whole, not as a sequence of conjoined and separable feet.  And, a really crucial point is that the metre and the rhythm are not the same, but can have quite a loose connection as indeed they do in iambic pentameter, where every line in a poem can be in the same metre but every line can have a different rhythm.  This is because the metre is not altered to fit the rhythm of the line by for example substituting feet, and this is because the metre is assumed to under-determine the rhythm: that is, only some aspects of the rhythm are controlled by the metre.

The Fabb and Halle book presents a particular theory of metre which is supposed to account for both English and the major metrical traditions of the world, many of which we discuss.  We take the view that in metrical poetry, the most important factor is that the line is fixed in length, with permitted variations: that is, a metre is basically a counting system, so that iambic pentameter lines are counted out to have ten syllables.  And if the metrical poem also has a rhythm, the rhythm is dependent on how the metrical elements are counted, which is always a counting by pairs or by triplets, which emerges as a rhythmic pattern.

Does 'no rhythm' mean lacking a patterned repetition of accent or stress?

 Or of syllable weight or tone type, yes. Most of the traditional Celtic prosodies have syllable counting but no regular pattern of accents.  French alexandrines control stress only very minimally. The North Indian arya metre groups syllabic units (matra) into sequences but doesn't pattern stress or weight.  Japanese poetry counts syllabic units but does not have rhythm. Metre doesn't always have rhythm but it always has counting.

Define stress.

Stress is a property of a syllable, and it is a psychological effect, not an acoustic one.  We judge the stress of a syllable in context, relative to adjacent syllables.  So stress is always relational, and not an inherent property of the syllable; Mark Liberman's demonstration of that was an important innovation.  Various acoustic properties can indicate that a syllable is stressed, such as if it is loud, or high, or long, or several of these together.  In words of more than one syllable, the stress pattern is determined by the generative rules by the relational mechanisms just noted, and is stable in the sense that you know where the stress is, it can be listed in a dictionary etc., and incidentally these fixed stresses in multisyllabic words are the stresses which are hardest to put into weak metrical positions in poetry.  One of the interesting things about stress in English and many other languages is that it is determined by counting syllables backwards from the end of the word.  It is a good example of how our knowledge of language is atemporal, because structure can be generated backwards.

What's the shortest possible explanation you can give for the fact that stress has been a fundamental aspect of verse form in English, assuming you agree that that's the case? 

I don't have an explanation, and it puzzles me. Why can't English verse more frequently use syllable counting with either no or limited control over rhythm, like the French alexandrin?  Or, to take another possibility, English syllables can be differentiated like Latin or Greek into heavy and light, and indeed syllablic stress is determined inside English words relative to syllable weight, but English has few poems in a quantitative metre and even they tend to use stress as well.  Perhaps it is just the dominant influence of some very major writers who wrote using stress-based metres.  Or there may be something about English sound structure which determines it, and some generative metrists do make these kinds of proposal about the relation between languages and metres.

Years ago in the Berryville Old Book Shop, in Berryville, Virginia, I found a reprint of an intriguing book, A Study of Metre by Thomas Stewart Omond (1903). Is this a book prosodists or linguists take seriously today? What do you think of Omond's effort to replace the 'crudity and dubeity of scansion by syllables' with 'scansion by time-spaces'? 

I cannot think of anyone referring to it.  There is a longstanding tradition of arguing that metre is organized relative to time, but no generative metrists argue for it; metres have declarative or map-like properties and not procedural or directional properties.  In contrast, songs are organized relative to time, and you can see the difference between the atemporal poetic metre and the temporal musical metre when you see how variable the ways are of putting the two together in a song based on a metrical poem.  Incidentally, in the Fabb and Halle theory we argue that iambic scansions are constructed backwards across the line, from end to beginning (while trochaic ones are constructed forwards), which shows again that at least in our theory metre is completely outside time.  Perhaps metrical poetry gains its aesthetic because metre is an atemporal structure which is matched to the temporal structure of speech in real time.

What would you like to know about language that you don't know?

I'd like to know whether poetic or literary language produces any kind of distinct epistemic feeling or arousal, in itself, just from its difference from ordinary language.  This is a view which to some extent reflects the traditional notion of 'tension' as a characteristic of aesthetic language and aesthetic objects more generally, but it's a notion which is very difficult to test.  I'd also like to know why in Ma'di it is not possible to end a past tense sentence on an object, for example why the translation of 'I read the book' is not a possible sentence but 'I read the book yesterday' or 'I read the book for sure' are both possible.  It's been puzzling me for almost 20 years, and I once pointed out to a group of bilingual Ma'di and English speakers that you could say it in English and not in Ma'di, and they completely agreed while saying that – as for many of the facts about our own language – they had never noticed; they knew it without knowing it.  I have not been able to find a good specific linguistic explanation for this, as a linguist should.

 

---

Above is a guest post by Mark Dow.

Update from Mark Dow (12/23/2022):

In our 2022 interview, Nigel Fabb and I briefly discussed T.S. Omond's A Study of Metre (1903) and its attempt to replace scansion by syllables with scansion by "time-spaces." I asked Fabb if Omond's book is much discussed by linguists, and he replied, "I cannot think of anyone referring to it."

I have since come across a discussion of Omond by Catherine Ing in her Elizabethan Lyrics: A Study of the Development of English Metres and their Relation to Poetic Effect (1951). Ing explains that Omond's "isochronous periods" are what we typically call feet, and she writes: "I believe that Omond's 'syllabic variety with temporal uniformity' is a true and illuminating suggestion if we interpret it elastically: if we substitute for 'uniformity' some word like 'balance' or 'proportion'; and if we look for this balance or proportion in units larger than that of the foot."

 

April 5, 2022 @ 7:29 am · Filed by Mark Liberman under Language and literature, Prosody

Permalink