oturn home > Secret ballet: the online version > About Secret Ballet (long version)

About Secret Ballet (long version)

'Secret Ballet' has been formed entirely from example sentences of the Pons Collins Cobuild English language dictionary edition of 1987 ('Helping learners with real english'), then a new type of dictionary developed by the English Department at the University of Birmingham. For the first time, words were illustrated by showing them embedded in examples of real usage. The examples were taken from the 'Bank of English': a huge online repository of texts ranging from journalism and technical writing to biography and fiction, and including transcribed recordings of actual spoken language. Telling by their flavour and content, most texts seem to originate from the Seventies and early Eighties; student revolts, women's liberation, political activism in general as well as many traces of the big machinery of the English political system are present in many example sentences.

When browsing the 1987 dictionary at length, one peculiar trait is the re-occurence of certain first names (Ralph, George, Jane, Felicity) and surnames (Castle, Hochstadt, Lithgow) in a fictional context. Apart from having identified 'Swallow' and 'Zapp' as fictional characters in a couple of novels by David Lodge (whose name also appears in an example sentence and consequently in 'Secret Ballet') and finding out about Diana Mitford and her link to Oswald Mosley via Google, I do not know to which novels or other sources these names belong (thousands are studiously listed in the front of the dictionary). However, it seemed clear enough that fragments of several longer texts are dispersed over the huge body of examples. This led to the idea that the whole dictionary could be seen as an exploded novel that reconstructed, would offer a material key to the Seventies and early Eighties. The many sources would be one merged source encompassing all aspects of society (with the exception of sex, to some extent): family life, profession, party politics, media, culture; crime, violence and madness; botanics and other sciences to the extent they have become more or less common knowledge; landscape, city life, tourism; cars and all sorts of things mechanic; friendship, hatred, jealousy and other emotions; childbirth, sickness and death. At the same time the tapestry of the things listed reflects all the ideological charges that creep into the very structure of language, intentionless. The example sentences are by no means neutral; they reflect the perceptions, preoccupations and prejudices of 'voices' of that era and thereby, portray its ideology in a total or stochastic way, eliminating the naturalising clusters of any partial discourse.

Comparing the 1987 edition with the current edition of 2002, it is surprising to note that hardly any sentence from the original edition has been retained (I could not find a single one when comparing both editions). One reason of course is that the dictionary concept favours up-to-date examples reflecting the most current use of terms. As the Bank of English' is constantly extended, more, and more recent, examples are brought up by clever algorithms as exemplifying predominant forms and contexts of usage. I got the general impression (this is hard to confirm empirically) that between 1987 and 2002, a move has taken place from using example sentences taken from the social and political sphere towards drawing on the personal and psychological sphere. While I entertained the suspicion that this might be due to changed politics of the dictionary authors or their publishers, I now believe it just reflects a profound shift in public conscience away from a broader political awareness towards an individualistic mode of thinking - a shift that is embodied in a quantitative sense in the Bank of English itself.

The text of 'Secret Ballet' has grown in the following fashion. At the very beginning, I began aggregating sentences around a few sentences that seemed to fit nicely. "A glow of light appeared over the sea" was among the very first. New sentences were found through aimless browsing, sometimes by looking up other terms from the same semantic context. At a later stage, this random or serendipitous approach was replaced by a linear reading of all example sentences, starting at the beginning of the dictionary. I carefully read sentence by sentence and, from letting each sentence resonate on my memory of the entire text, more or less immediately identified the context where it might best fit. At this position I tried to insert it, if and where it may fit grammatically. If a sentence failed to resonate, I moved on, leaving it for the (hypothetical) next cycle through the dictionary once 'Z' has been completed.

As the text grew, the number of potential sentence insertion points also increased, making it possible to accommodate an increasingly higher percentage of sentences. This also brought back conscious design choice to some extent, since more than one part of the text might resonate and offer itself for insertion. At the same time, it got harder to summon the entire text in memory to determine the best fit. Parts of the text would be visited infrequently, beginning to dry up or showing signs of neglect. Some of theses parts were removed or split up and relocated in the last phase of compilation.

During the last phase of compilation, I tried to 'round off' the text, mainly in order to close narrative gaps. (This implied that by then I had given up the hope that I might be able to complete the work in line with the original concept. Having reached letter 'C' after a couple of years, I estimated that it would take the rest of my life to finish the whole alphabet, let alone going into the second recursion to incoorporate sentences left out in the first pass.) I therefore moved back towards a more directed search for fitting example sentences, often looking up synonyms or words belonging to the same semantic cloud. This turned out to be a very frustrating and time-intensive process.

As described in the note preceding the text, I allowed for changes of first names (sex remaining unchanged) and of surnames (but disallowing a change of surname for first name or vice versa); flipping of sex in personal pronouns ("he asked her out" can become "she asked him out" but not "he asked him out") and simple concatenation where possible (the 2 sentences / fragments "a man in his fifties" and "Castle wore a grey suit" can be combined to "a man in his fifties, Castle wore a grey suit".

The formal aim during compilation was to create a text as smooth and 'natural' as possible; the different origins of the constituting sentences and the friction created where they meet seemed to introduce enough ambiguity. The plot seem to come about by itself, although it is obvious that the results of resonance (producing the place where a sentence seems to fit) were in an oblique way related to my own personal, social, and psychological setup. Intentionality was effectively prevented in that it was close to impossible to ever find the sentence that *I* would want were I an author of fiction in the traditional sense (or only at the price that the procress of writing would slow down to snail's speed).

Aiming for 'smooth narration' with varied and disparate sentence material from a multitude of sources must fail, but this failure constitutes the form of the work. Looking at each sentence in isolation, other contexts than the one of its given position appear; however, the pull of narration seems to drown this fragmentary 'Eigensinn' (proper sense) in an often clichéd and at the same time erratic and unevenly paced plot. The endemic shift of aims and intentions, a certain feel of characters changing inexplicably or even appearing spineless here and overacting there, are part and parcel of the method and may be regarded as its stylistic core.

I have posted another text on secret ballet, an interview by Tay Arrow Sherman of one38.org that was never published (for reasons I can only speculate about).

D.W. (who prefers to remain anonymous) has contributed an interesting comment on secret ballet, including some references to related approaches, to which I have replied. And there is another comment by Joe Gone of melting object.

Go to Index of Secret Ballet

Last update: 16 October 2008 | Impressum—Imprint