Abstract: An atttempt to demonstrate how and why cultural change occurs continuously and independently of any social dynamics and simply as the result of the working of ordinary cognitive processes understood here without recourse to specific versions of cognitive theory.
The past is a lost world. We know so little about it. Of all the trillion upon trillion conversations, speeches, sermons, invocations, recitations, chants, songs, concert performances, which took place before the 1870s (at the earliest) hardly a trace remains. Their sound is recorded nowhere; the memory traces which they left are gone because everyone born before the 1870s is dead. Only where something was written down in a musical score, the text of a play or a prayer, or an entry in a diary, does something remain. The farther back in time we go, the more reconstructive (and hence at least partly speculative) is any attempt to reproduce the sounds and gestures of the past. We can reconstruct the conversation at a nineteenth century country house dinner table guided by what we read in Jane Austen; an eighteenth century concert fit for a king guided by a musical score; a seventeenth century theatrical performance guided by all that Shakespeare researchers can tell us; a sixteenth century speech by Queen Elizabeth the First as handed down in the history books …. but in the end only one thing is certain, that somewhere we will have got it wrong. We are guessing – we are theorising and we are improvising. There are no Big Data such as we now possess in sound and visual recordings and even those, I will later argue, are fraught with problems.
I am making an important assumption. Surely, you might say, there is at least a possibility that person B who heard it from person A passed it to person C who then passed it on to person D, and so on down the line so that a twenty first century rendering of a prayer may be sounded out in pretty much an identical manner to the way it was sounded out ten or twenty generations ago. I don’t believe this is true and rather than just say Haven’t you heard of Chinese Whispers? I want to use this essay to argue that case.
In logic, if A implies B, and B implies C, and C implies D … all the way through to Y implies Z, then it follows that A implies Z. Logical implication and entailment just is a transitive thing.
But if I understood my mother when she talked to me, and she understood her mother when she talked to her, and that mother understood her own mother …. it does not follow that I would understand my 26th great grandmother if she were still alive to talk to me. Indeed, it is at least very likely that I would not understand. There will be a failure of transitivity.
Such failures arise in two main ways. First, the things I want to speak about are different to those my 26th great grandmother wanted to speak about. My vocabulary is different, full of words which would be to her incomprehensible neologisms. Likewise, her vocabulary included words which have fallen completely out of use and which can now be found, if at all, only in specialist dictionaries. It is a perfectly general truth that the world we inhabit change and the words we use to talk about our world change with them, sometimes very rapidly, sometimes more slowly.
Second, the way we speak changes over time under both external and internal pressures.
Pronunciation, accent, intonation patterns, all change – indeed, are changing all the time, never stop.
These simple ideas can be given expression in terms of very simple set theory. There is a set (almost certainly fuzzy at the edges) of all the possible utterances which I can passively understand should they be addressed to me. That set changes over a lifetime – in my late teens I could understand utterances in Swedish because I had learnt some Swedish but which I would no longer understand because I have forgotten what I learnt – but importantly the set includes most (perhaps all) of the utterances which my mother ever addressed to me. But my set does not perfectly overlap with her set, not least because the world has changed a great deal and now includes the internet, emails, and so on indefinitely. Included in her set were most (perhaps all) of the utterances which her mother addressed to her, none of which were ever addressed to me because this grandmother was dead before I was born. Probably there were things which my grandmother said which I would not understand if by some extraordinary means I could hear them now.
I can even give a sort of proof. Recently, going through things which belonged to this grandmother, I found the printed prayer which stood like a photo in a frame on her bedroom dressing table. I looked at the back where I found printed the words MOWBRAYS’ DEVOTIONAL GLAZETTES G 7. Well, I know that Mowbray is a big religious publisher confirmed by the later words A.R.MOWBRAY & CO Ltd. London & Oxford – something which has not stopped the company putting the apostrophe in the wrong place, useful evidence that the decline of civilisation did not begin last week. And I know the word devotional. But what about this glazette? I’ve never come across the word before. It sounds like it has some connection to gazette, but what connection? So I google and for the first time ever, Google really struggles. There is no definition of the word on offer anywhere but there are a handful of other uses which Google finds on ebay, including uses to brand-name early (1890s – 1910s) picture postcards which have a glazed surface – what we would now call a laminated surface. So here we have a word which my grandmother would have presumably known and understood but which, unaided, I did not. And here we are only talking about the very, very recent past.
It takes only a bit of imagination to see that over time, the sets of all possible utterances which a person in generation Z can understand have migrated so far away from those which a person in generation A could understand that there is eventually minimal or no overlap at all. There is a more or less complete failure of transitivity. Or, to put it in the language of set theory, sets migrate. If this is true of language, then it will be true of many other forms of expression. That claim requires some fleshing out.
Why is it not possible for a singer in an oral culture to orally transmit a definitive version of a song to an apprentice singer who then in turn passes it on to the next generation apprentice, and so on, indefinitely?
First, and not at all trivially, a singer may not care at all about a definitive version and from performance to performance may vary in all kinds of ways the song they sing. The singer acts creatively, improvises, but is also affected by how much they have had to drink, how much they like the audience, and so on through an indefinite range of possibilities. So an apprentice has to somehow figure out what is essential to learning and reproducing “the song” and what is incidental. And there is no guarantee at all that all apprentices will figure out in the same way, even if they have never heard the expression “cover version”.
This is really a way of introducing the idea that in relation to cultural transmission or reproduction, there is always and inevitably an inescapable moment of interpretation. That idea can also be expressed in the claim which tells us that a theory (an interpretation) is always underdetermined by the data which support it. There is always more than one way to skin a cat.
Second, it is extremely rare for something like a song or a dance to have only a unique performer at any one moment in time and for that unique performer to have a unique apprentice. At any one time, the performances of a living group of performers are attended to by a group of apprentices. When a culture is dying, one way of showing that is to point to the fact that the group of apprentices is smaller than the group of established performers. But dying or expanding, the performances of current performers constitute a set (almost certainly fuzzy at the edges) of what constitutes the empirical reality of a particular song. When a couple of centuries ago (or less), ethnographers began to collect the words of folk songs one of the first things they had to cope with was the huge variation between versions of what were, in some sense, renderings of the same song. The set which made up a song was not only fuzzy; it was positively indeterminate.
I once encountered a real-world near-demonstration of this truth. In the eighteenth and nineteenth century, men from the Orkney Islands off the north coast of Scotland found employment with the Hudson’s Bay Company in Canada which traded with the Inuit population of the north. Some Orcadians also served on whalers which hunted in the Canadian north. These men left behind not just babies but fiddles and the knowledge of how to play them. Long after the Orcadians left, there continued to be Inuit fiddlers playing Orcadian music. In 1978, holidaying in Orkney, I found myself one evening attending a unique, first-time-ever gathering: a group of Inuit fiddlers had travelled to Orkney and they were going to play their fiddles. Then some Orcadian fiddlers were going to play theirs. Then, it was hoped, the two sets of fiddlers would play together. Well, they tried and they partially succeeded but you would not have called it a successful jam session. Cultural drift had taken Inuit and Orcadian fiddlers in separate directions and after a period which could have been more than 150 years, they were playing differently from each other. But in between, sons had learnt from fathers in unbroken chains. The line of descent was there but the music had migrated enough to create quite a lot of intransitivity.
But suppose there is just one performer and one apprentice, and suppose the performer is an obsessive about maintaining just one version of a song, a dance, a prayer and boxes the apprentice’s ears when they deviate ever so slightly from the standard? Surely this can stop the set from migrating, stop later versions from becoming intransitive with earlier ones.
I have six inter-connected arguments to offer against this possibility. One I develop through the concepts of foreground and background. That is connected to what I call the sample book problem. Then there is the important problem of forgetfulness. Then there is a problem of context which is often overlooked. There is a problem of finite intelligence. And, finally, there is the problem of unique experience.
Human beings, constructed as they are, can only attend to some things at any one time out of all the possible things which could be attended to at that moment. They can, in fact, only notice some things out of all the possible things that could be noticed. About these things to which they attend or which they notice, they can say at least something, though even then it may be sketchy and not very informative. Nonetheless, it is a whole lot more than can be said about all that was not attended to, was not noticed.
There is a foreground of experience - things we notice - and a background of things we don’t. Quite often, we can change our focus of attention in order to pull something out of the background and into the foreground. But not as often as we may imagine and quite often we only do it when prompted.
This is not just a fact or truth about perception. It is perfectly general. Everything we experience and everything we do is handled within the frame of foreground and background and there is no avoiding that fact. The consequences are multiple, not just for everyday life but also for things like artistic creation and political understanding.
The contrast of foreground and background is played out very obviously in relation to spoken and written language. Parents, teachers and ministers of education habitually foreground some bits of language as particularly important for children to learn. They emphasise bits of pronunciation, bits of grammar, bits of punctuation as things which are very important to get right. Quite often, these foregrounded features are selected with a view to stopping some incipient change occurring. They are conservative measures.
You can only practise so much vigilance. No ordinary parent or teacher or even minister of education can be googling all the time. As a result, some changes get past even the most vigilant defenders who are thus always in the position of King Canute, unable to turn back the waves. Change will happen even on those language fronts you have opted to foreground and defend.
Worse, there is the background of language use still to consider, all those things which aren’t being attended to. Here change happens unnoticed even by those who pioneer the changes. They just do it without knowing that they are doing it or why they are doing it. I will give an example from written language. Recently, I was reviewing articles and chapters I had written forty or fifty years ago. As I turned some of them into new Word documents, I realised there were things in them that I now write differently. Decades ago, I would have written U.S.S.R. and N.A.T.O. and B.B.C. and so on indefinitely. But nowadays I don’t do that. I write USSR, NATO and BBC and so on indefinitely. But there was no point at which I was conscious of dropping the stops and I did not know that I had indeed made this discontinuous change in the way I type until I got involved in reading my own old work. Likewise, there was that point when – like many people –I stopped speaking and writing Roumania or Rumania and switched to Romania. Don’t ask me when or why.
So in the background, even though it involves things which we do quite consciously – as when we sit down to type – changes happen which are not reflected on or brought into consciousness at the time they occur and may indeed not be noticed until much later. Historical linguists come into their own studying such changes but what is happening here is not specific to or confined to language. All things change, whether we like it or not or whether we notice it or not. And All does mean All, even those things we may imagine are under our own control. To return to language, if you think there is something called The Queen’s English or BBC English which does not change, just try listening to a fifty year old recording of a BBC radio broadcast or a fifty year old Queen’s Speech.
So far, the argument amounts to this. Everything we experience and think about is handled in terms of foreground and background. We have a bit more control over foreground but not enough to prevent change even when we are trying to prevent change. We have less control over background, often none at all, and background changes all the time and sometimes very fast. Change goes on in the background willy-nilly, as I have already suggested with the example of pronunciation. We are capable of changing things in the foreground and liable to change things left in the background. The dynamics of change are different in the two cases.
In the foreground, we are liable to influence from others (often massively so) but when we make a change it is often (perhaps always) of a discontinuous nature and involves a decision on our part, as when we decide to quit smoking and move from being a smoker to being a non-smoker which in turn makes us an agent within a broader cultural shift. But when something in the background changes, unknown to us, it often does so in a way which has a sort of continuous character. So we start to say “Hi” instead of “Hello” on a few occasions, without knowing why we pick those occasions, and then we start to use “Hi” more often and, perhaps eventually we move to a situation where we become a monoglot “Hi” user rather than a monoglot “Hello” user. But we didn’t decide to do this. We still understand what other people mean when they say “Hello” but we just stop using “Hello” ourselves.
(The contrast between continuous and discontinuous change is important. There is a long history of theorising about the contrast, with the science of geology having been a major site for the early discussion. Nowadays we are most familiar with the idea from the way we contrast analog with digital. Think of clocks).
How does all this apply to the single performer with a single apprentice? Suppose it is a singer and a song. The singer has a unique voice profile (as modern technology knows) and it changes through time: an old man does not sound like his younger self. The singer can’t do much about this and almost certainly discounts it and consigns it to the background when teaching an almost certainly younger apprentice. The singer can only foreground so much of the song and its singing and inevitably some things will pass unnoticed. Maybe the singer takes four minutes thirty three seconds (on average) to sing the song and the apprentice takes four minutes thirty one seconds. If you don’t notice and stop that, then the song has already changed. In contrast, when the singer foregrounds something, like a drawn out note or word, then that does mean that the apprentice may well get their ears boxed for getting it wrong.
There is still a double problem. The poor apprentice has to understand what they have got wrong and find a way to correct it. Because of the ubiquity of the need for interpretation, the apprentice has first to correctly identify what the singer is so agitated about. The problem is analogous to that children have when their speech is corrected and they have to grasp what it is that it is being corrected. Not so long ago, I listened to a young child reciting numbers from one to twenty with complete accuracy. But at the end, his father intervened to say No, not twen-ee, it’s twen-tee. The child was completely baffled by this piece of linguistic correction which had absolutely no connection to the task he had set himself of reciting the numbers in correct order. How was the child supposed to know that though living in south east London he was not supposed to speak like south east London?
Even if he had been a budding theorist of cultural arbitraries, there would still be the problem of converting advice into successful practice. Children do often get it right in the end, though it probably has little to do with advice they are given, and as anyone who has ever learnt to drive a car will know, giving advice is more easy to offer than to act upon.
Worse is to come. In oral cultures, singers forget today what they prescribed yesterday or, perhaps to make it more plausible, they forget next year what they prescribed this year. They have no sample book outside their own heads and we all know that our memories are constantly re-organising themselves. They have no means of comparing the sample which occurs to them today with the sample which they were using yesterday, let alone last year and which is almost certainly completely forgotten.
But suppose there is a real sample book in the form of a voice recording, even a film which shows all the accompanying gestures and so on?
The very same problems recur even if they seem less severe. Eliminate the possibility that the singer says they were having an off-day when the recording was made, there is still a problem of determining what is foreground and what is background in the recording, what matters and what does not matter. If the singer looks up at a certain point, does that matter or is it just because a bird was flying overhead at that moment? Then again the apprentice has to convert what is available in the recording into a new performance which uses the apprentice’s unique voice rather than that of the recorded singer, and so on.
Sample books do not solve any problem in some automatic way; they have to be interpreted and a regress can only be stopped by making a decision: this is the way we will do it. And decisions, one might say, are fatal to the integrity of cultural transmission. The decision indicates what will be allowed to Pass and what will Fail. But on a different day, or with a different judge, it is entirely likely that the bar would have been set higher or lower.
Imagine the teacher listening to the apprentice and eventually declaring That’s it! or perhaps merely That will do! That’s a decision, not something completely grounded in the sample in the teacher’s head. Nor is there any guarantee that next time around the teacher will come down in favour of the same version; the teacher is capable of forgetting the sample used last time and also capable of unconsciously modifying it. The mind is always at work, in one way or another, and it is the mind at work which makes all culture unstable even in what we might think of as an otherwise unchanging world.
There is a further peculiar, problem created by the fact that all our activities have a broader context.
Background and foreground are separated by temporary boundaries – things move in and out of focus, the change triggered for many different reasons. But even when something stays for a long time in background, even deep background, it exerts an influence on what goes on in foreground. I use a hypothetical example to develop the argument.
Imagine a culture in which it is expected that certain utterances will be produced in a voice which is loud, clear and decisive. Maybe when a prayer is spoken or a sentence handed down by a judge. But suppose that in the wider culture there is an unmonitored and untheorised drift towards quieter forms of speech. The explanations could be purely external and chance. Maybe people are living in a police state and fall into the habit of talking in whispers; maybe more and more people work in open plan offices or live in apartment blocks with flimsy party walls. Whatever, people are talking more quietly. In this situation, the priest or the judge who continues in the old way will begin to sound ridiculously loud rather than impressively loud. Quite unconsciously, but affected by what is happening all around, an officiant shifts towards dropping their voice by a decibel or two. Should their audience contain an old-school office holder in retirement, that person may be saying to themselves Speak up! Speak up! because they happen to be outside the loop of an ongoing, broader cultural change.
It is in such continuous contextual interaction that I think we may find part of the explanation of cultural changes which it seems no one intended but which have happened anyway. For example, if a broader culture gravitates towards greater informality of style, then that may provide a kind of push towards making things like weddings and funerals more informal, even though those are things which most people might be happy to regard as governed by tradition and to be kept going in their older forms. Context is not a sinister force, but it is a powerful one.
There is one final and very important thing. Our brains aren’t big enough and the time available to us is so short that it’s not possible for every bit of cultural material to be given foreground attention. That implies that those who want to stop cultural change cannot win every battle because they can never have enough troops to deploy. There aren’t enough hours in the day for anyone to stay on high reflexive alert to more than a small number of things which may change if not attended to.
To write English properly, you are supposed to master apostrophe rules. As it happens, they have a rather complex and confused structure which make them very difficult to learn without a quite disproportionate expenditure of effort. Very few people master this glass bead game. In this case, there is a long-term dislocation between a set of rules which tell you what you are supposed to do and what is actually done. The most likely resolution is that the rules will eventually be abandoned.
The problems which arise from limitations of time and intelligence can be seen in comic form in the desperate, expensive and futile attempts which English schools make to get pupils to conform to school uniform rules. The Deputy Head goes on offensive against jewellery but misses what is happening to finger nails; they switch to finger nails and miss what is happening to skirt hems; they focus on skirt hems …. The only sound conclusion available is that they never will succeed because they never can. To a disbelieving audience, King Knut proved that claim a very long time ago.
Last but not least, people do not share the same experience set, the set of things which happen to them and which provide the raw material for their minds to work over, interpret and act upon. Experience sets are unique to individuals. Every day and all over the world, many millions (maybe more) people use or hear used the word Heathrow but only a very strange fluke would ensure that over time they have identical sets of Heathrow experiences. In all probability, they hear the word pronounced in different ways and out of that experience they have to fashion their own pronunciation, much affected by the language context from which they are working – Cantonese, French, Russian…. Very few will head to the online forum where such things are discussed and even then the effect will not be decisive.
We do not have a Big Data set which harvests the sounds of each day’s token utterances of the Heathrow type. There is no central depository, only the experiences of millions of individuals. A linguist with a sample of all the utterances will be able to sort them into sub-types – for example, the sub-type HEATH-row with stress on the first syllable and the sub-type Heath-ROW with stress on the second. The linguist may be able to hypothesise that the distribution of sub-types have shifted over time, the first pronunciation (American) overtaking the second and original English pronunciation for reasons much connected to patterns of global aviation. So we have the beginnings of an account of cultural change. The only sure thing is that it would be absurd to suppose that the continuously updated pie chart breaking down Heathrow pronunciations into their variant forms could have remained unchanged even over the short period of time in which the airport has existed.
The lines of argument developed in this essay apply equally to the understanding of changes in beliefs, belief systems, ideologies. Those who wrote the religious texts on which many cultures have relied probably thought that they were settling things for the future. In fact, as everyone knows, they simply provided data for an indefinite number of ever-changing interpretations. The human mind seems to like nothing better than the challenge of a text. This inherent instability in what in some cases are presented as unchangeable belief systems has one major advantage. It also allows for scientific progress and revolution. Of course, there is also an external dynamic provided by migration, war, conquest and economic change. But even without that external dynamic to prompt it, human minds are always churning.
© Trevor Pateman 2018; first published in slightly different form on my Blog readingthisbook.com, September 2018