I remembered recently the existence of a letter my stepfather, the British poet and novelist Douglas Oliver, wrote me thirteen years ago to explain, on my request, the series of experiments he conducted in his study of prosody and voicing, Poetry and Narrative in Performance. The book was published in 1989, and I think the recordings that he describes in the letter and the subsequent analyses (very densely related in the book) must have taken place a few years earlier. I’m very interested in the matters discussed in the letter, and as it will have been ten years this coming April since he died, Doug is very much on my mind. But the work he did is the point, and the focus of my attention, so I’d like to share this letter. The length of the letter necessitates it being divided into at least two posts. Doug is writing from Paris; I am 24 and living in San Francisco. To a very tiny extent the language and tone of the letter is pitched specifically to me, but I think it is by and large available to any interested reader:

Tue, July 2, 1996

Dear Anselm

You asked me to describe the basic themes of Poetry and Narrative in Performance: it’s a very dense, technical book, so I’m just going to describe those themes which relate to poetry, not to fiction.

My central concentration is upon the idea that a poem has a possible infinity of meanings depending upon the individual response of readers. Allied to that, does it make any sense to say that a poem has a particular music natural to it – since, again, a multitude of readers when reading it give it a different music? Is it elitist to say that any reader’s version of the poem is superior to anyone else’s and is even the poet’s own version subject to this? Such an ideology fits in with all manner of other kinds of philosophy which are currently fashionable: multi-culturalism with its insistence that no-one’s culture can be challenged without taking up an elitist or power-driven position; anti-foundationalist philosophy, which states that there are no truths external to language and to our individual expressions of them: that both truths and the “self” therefore are social constructions and have no warrant outside language.

This line of philosophy starts with Wittgenstein, and runs through Heidegger, Derrida, to people like Richard Rorty and Stanley Fish. Rorty, in particular, has promoted a new form of pragmatic philosophy which ran very hot at the time I wrote the book – and even today the new philosophy books bought by the American Library in Paris carry that stamp: they’re about Nietzche, Dewey, Rorty, and “anti-foundationalism”.

I’m not reactionary about this: I believe 100% in the multi-cultural attitude and almost entirely (with one vital reservation) in the notion that truth is a socially constructed entity which is difficult, perhaps impossible, to extricate from its imprisonment in language. However, if no one reading of a poem is better than any other reading (an obvious nonsense), then all my own readings of my own poems are as good as any other readings! And if that were true, I could never improve either my reading or my writing of my own poems. And if a given reading is superior there would have to be, I think, some generally shared notion of what a good reading consists in. The idea of an external standard of truth comes very near us then, though it is not quite reached. I am more interested in the fact that we nearly reach it than in the fact that we don’t finally do reach it; and that is a difference between me and the anti-foundationalists.

So I have this reservation: there is a crucial distinction between describing an experience and performing it. There is a similar distinction between trying to fix Truth into a single description (which would be wrong) and half-sensing a perfect truthfulness as a possibility hidden within our actions (performances). The first kind of Truth would be dogmatic and religious; the second kind is a non-existent entity which nevertheless seems to guide us – a real mystery in fact.

If I describe my “self” I see it as a social entity: if I “perform” my “self” by being it, it is something more mysterious. Similarly, if I try to describe the tune of a poem I can’t without seeking a consensus in society about what tune it should be. If I perform the poem, then, for that moment, there is only one tune that I’m trying to make. The description is public and subject to all the problems of truth as socially constructed. The performance is private, interior, and almost indescribable. In poetry, by “performance” I mean that moment when the poem is first written down (created), or read out loud by a reader or “performed” silently in the reader’s head.

What I have done is to get different readers to record performances of the same poem by reading into linguistic machinery with electrodes round their necks. The results basically give a graph of:

  1. the rise and fall of the voice (the intonation)
  2. plus the speed at which the sounds travel (duration)
  3. plus the presence of any pauses in the reading
  4. plus the patterns created by those stretches of sound when the larynx (voice box, adam’s apple) is continuously sounding (during the speaking of vowels and voiced consonants) and those other shorter moments when unvoiced consonants occur. In the following words, I have underlined the voiced parts when the voice-box is sounding:

buzzing wasp

Try it: b and z make your voicebox vibrate, w is made only with
the lips and air, and so are sp.

The significance of point 4 is that we are able to make the tune of a song or a poem only when the voice box is sounding: it is in the throat that we make the fundamental frequency of the voice, and it is in higher parts of the vocal apparatus that we make all the higher frequencies. The ear chooses to regard only the fundamental frequency as the tune. This has been much neglected in the study of poetry, in my view. Stretches of continuous voicing affect the pace of a poem and also such questions as the continuousness of a held thought or awareness. My favourite line to show this is from Wyatt:

So unwarily was never no man caught
With steadfast look upon a goodly face…

where the whole of that first segment shows up as continuously voiced and helps to convey an expression of continuous rapture at the sight of the beautiful face before the word caught raps at the end of the line to catch our attention back again. (The w, by the way, is strictly unvoiced but the nasalization of the n can just about maintain the voicing while the lips are forming the w.) The role of voicing in a poem is never talked about by anyone, but I believe I have shown it is complex in its effects.

So I’ve made all these recordings and compared the results from the different readers. They were asked not to read “dramatically” but to feel for the neutral music of the words. There is, in fact, some old work by a German linguist, Sievers, showing that it is possible to identify the neutral music.

Then, in some of my experiments especially, I have asked audiences of, say 20-30 people, to identify the “best” readings, so that I can escape, as much as possible, my own subjective, “elitist” judgments about which are good and which are bad readings. This is the social consensus I’ve talked of.

I have then developed a complicated method for comparing the graphs of these “best” readers and measuring them against those of “worst” readers. The basic results are these:

(a) It is important first of all to make sure there is a broad consensus
about how to interpret the meaning of a poem

(b) If a poem is orthodox metrically, inexperienced readers generally find it
easier to decide what tune to give it.

(c) When performing orthodox, clear poems, the “best” readers tend to
create markedly similar tunes. The “worst” readers typically read
with a flat intonation or give a unorthodox interpretation
of meaning. This does not alter (c).

(d) Talented alternative versions of the tune are possible, but will be
perceived as “dramatic” or unusual in some allied way.

(e) If a poem is very experimental in its prosody, inexperienced readers
may mess up in reading it. Again, this does not alter the convincing
evidence from (c) that it does make sense to talk of a neutral tune for
a poem, providing the reader knows how to interpret it.

People sometimes think I’m being Platonic: claiming that there is a perfect tune that arises within the performance like an Ideal form. In fact, I believe that this perfect tune is ineffable, which is the same as saying it doesn’t quite exist. Nevertheless we sense it as a possibility.

Originally Published: October 30th, 2009

The son of poets Alice Notley and the late Ted Berrigan and stepson of poet Douglas Oliver, Anselm Berrigan earned a BA from SUNY Buffalo and an MFA from Brooklyn College. His collections of poetry include Integrity & Dramatic Life (1999), Zero Star Hotel (2002), Some Notes on My Programming...

  1. October 30, 2009
     Joseph Donahue

    Thank you for posting this. Looking forward to the rest.

  2. October 31, 2009

    Dear Anselm,\r

    Thanks for posting this fascinating series of insights and thoughts. The distinction between performance and description sparks lights of recognition; and the discussion of voiced vowels and consonants, with the quote from Wyatt, is terrific.\r

    Sometimes I think about a poetry consisting entirely of voiced consonants,\r


    or an alternation of voiced consonants with their unvoiced twins,\r


    . . . but I hadn't thought about the role of voiced consonants in lyricism. \r

    p.s. The hint of a personal address in your step-father's letter is entirely conventional and lends itself to public discourse; you're right to think that the letter is available to any interested reader, without hint of gossip. Thanks again.

  3. October 31, 2009

    Interesting analysis, what will take me more than a couple of readings to get. Immediately, however, the package put me in mind of Old English, or Anglo-Saxon, prosodic rules, what held the field before the Norman conquest brought with it latinate rules involving syllable quantifications. Trying to figure out why I pull down my trusty Princeton Encylopedia, go to the rather lengthy entry on Old English prosody and who should I find referenced in the context of the same but your 19th C German linguist, Sievers.\r

    Looking forward to part 2.\r