Friday, December 7, 2012

Contexts, Reading and Comprehension, pt 1

I realize most readers here are genuinely literate. By that I do not just mean the ability to see a written word on a piece of paper, a screen or a metal plate and realize that the word there is, indeed, 'STOP', or whichever word it happens to be, but the ability to reconstruct in their minds the thoughts expressed in a text and after that, think a thought sufficiently close to the one the author expressed that for all purposes and intents there is a family resemblance between the thoughts of the author and the reader. Now, if an author is bad at expressing themselves, it may be impossible to know what the intended message was.

An example of this would be this phrase, which has been turned into an internet meme:

Has anyone really been far even as decided to use even go want to do look more like?

Even given the context in which it was posted, it is difficult to figure out the thought the original author wants to express. We know this much: the thought is a question, and given that we know the specific form of question used, we can be fairly sure it is supposed to be answered either by yes or no, we may even surmise that if the answer were yes, the person asking would want to know the examples that do satisfy the condition. I am informed the actual context suggests the meaning is something like has anyone ever decided to do more in order to produce realistic graphics (than this particular game's producers have)?

As readers, we must be entitled to assume writers master the language in which they are writing - that if an author writes "kill all communists like vermin", he has not been mislead by a malicious teacher into thinking kill means love, all means your and communists means neighbour, like means like and vermin means yourself. Getting the meaning of words right does not necessarily mean they will communicate their thoughts perfectly - there may very well be occasions when slips of the pen (or of the mind) obfuscate the intended idea even if the vocabulary is right, and there may be occasions where mistaken vocabulary still is rescued by the context in which it appears.

 However, having read the "rebuttals" of my criticism, I am increasingly disheartened. It seems my main detractors to some extent refuse to think in terms of reasonable communication. So I figure I have to write a thing about communication.

Communicating an idea

Our brains are neural networks. Neural networks are finicky things, in some sense. They are good pattern matchers, but they also make mistakes, identifying so called false positives. This can be easily observed on occasion: talk a bit too fast about a concept, and it is likely your hearers will get the wrong ideas, they will find some pattern either along the lines of what the listener often thinks of, talks of and is spoken to about. Communication is not perfect, but for this reason, our use of language contains strategies to improve the chance of being understood correctly.

Here, it is reasonable to mention the Gricean maxims. In short, these maxims give guidelines that both parties of an utterance - speaker as well as listeners - benefit from. We use them both to form our utterance in a way that is easy for the recipient to parse, and to parse the utterance assuming it is formed along similar guidelines. 

I will try and describe my understanding of how these maxims apply to writing and communication more generally. I would love to hear objections, especially if Acharya's fans are capable of providing reasonable justifications for some of their claims.

The Gricean maxims are the following:
  1. say truthful things
  2. say relevant things
  3. say a suitable amount
  4. say things in a clear manner
There are of course situations when we will violate any one of these, we lie, we ramble, we fail to be clear. A good speaker or author, however, will at least get the three last ones right, and an honest good speaker adheres to the first one as well.

There are various sources of failure, though. For #1, we may have false information available to us or be under the wrong impression about how things are or we may tell an intentional lie. Epistemology can cause any number of problems, and so we can of course only say things that are true as far as we can tell. When reading a scientific text, we assume the author thinks his claims are true. However, we are permitted to suspect his experimental readings may be due to any kind of mistake in the setup of the experiment.

As for number #2, this is difficult. We may ourselves have an unclear idea as to the relevancy of a thing in the wider context. However, if we make a statement, and this statement consists of smaller statements, this should indicate that we believe all the smaller statements to be relevant to the full statement. The reader or listener should not have to look for the point we are making by having to ignore the right bits of what we say. If I were to tell someone how to cook a dish, and I told them:
  • cut about a third of a cucumber in short, thin sticks with a knife
  • put the sticks in salt and a couple of tablespoons vinegar for about half an hour in a small bowl or large enough cup
  • whip up a pancake batter with 3 dl milk, 2 eggs, 2 dl flour and a little salt, which you whip together in a bowl
  • fry four to six small perch fillets in butter in a frying pan until they get nice coloration, which is slightly brown
  • cook potatoes in a saucepan or other suitable thing, on the stove
  • pour the cucumber-vinegar-salt mixture over the perch, let sit on low heat for about five minutes with a lid on still keeping the perch fillets in the frying pan, now you can put the small bowl or large cup in the dish washer, or in the sink if you don't have a dish washer
  • turn on the oven, 250 degrees celsius should should be suitable
  • pour about a deciliter cream over the perch fillets and let sit on even lower heat if possible for a few minutes without the lid on
  • Serve the fish and cream-vinegar-cucumber sauce with either mashed or whole potatoes on plates on the table around which you keep chairs for the guests or family members to sit on while eating, and maybe a simple but fresh salad on the side but make sure the salad's not wilted
Did I do a good job explaining how to make the dish?
(The exact amounts of salt and vinegar and cream may be wrong here, gonna check them tomorrow. Pangasius or pike-perch are also acceptable substitutes if perch is not available to you. It is a lovely little dish.)
I would say I did not do a good job explaining it. I violate a few maxims here!

Notice how there are two entirely irrelevant bits in the recipe? (The other weird thing about it will be explained later.) The reader who wants to make this dish would have to notice the irrelevance of two instructions - this is not a thing you normally expect when reading a recipe. If you are used to cooking, you are very likely to notice there are superfluous and irrelevant instructions in the recipe, but if you are new to cooking and have no idea whatsoever what is going on, you may actually turn on the oven or whip up a pancake batter while also carrying out the proper given instructions. Unnecessary details like these do occasionally creep into academic writing or other instructions, but generally these are of the type "go have a cup of coffee while waiting for the reaction to finish" or somesuch, things where the reader can be assumed to know that having a cup of coffee does not affect the result, and that the coffee instruction simply signifies that some time will pass when you simply will have to wait for the results. In such situations, I would wager they serve some kind of sociological role as well: indicating the widespread acceptance of such behaviors within the relevant subculture. It also is related to the phenomenon we call humor, and in the case of English it may in part have to do with avoidance of repetition.

Saying a suitable amount in part overlaps with the previous maxim - but both maxims have bits that do not overlap. The above example would violate this maxim as well, and the violations are marked by italicization. Anyone who has fried perch knows it never turns teal or blue or violet and therefore the detail about the color it is supposed to turn is superfluous. Certainly, it is not a misleading detail, but it is unnecessary. These kinds of superfluous details can be introduced for humorous effect, but generally that is avoided in academic discourse unless made very obvious. Excessive detail can be helpful for someone entirely new to cooking - but what is excessive or not depends on the intended audience. Certainly we could write over-specific texts about anything, and at that point it becomes a muddle of needless - even if true - details. We could go into detail regarding the number of legs the average table has, the variety of table designs, the historical development of the table, the physics of how much weight it can carry and still be entirely truthful. This would violate the maxims of relevance as well as quantity. 

If too little is provided, another problem appears: the reader or listener is assumed to know more than he or she does know. The reader will fail to understand the underlying thought. If, in an essay or paper or blog-post, I say "So-and-so theorizes that ..., and this reasoning seems compelling enough. It further fits the evidence provided by this and that" and what I really mean is "So-and-so is dead wrong on this issue" or even "So and so thinks this, and I have no stance whatsoever", I have not provided enough information - the theory of So-and-so can be relevant, its qualities might be relevant, the extra evidence that makes it look compelling may be relevant, but if I still do not accept it, I should say so in a clear manner - unless I have made this clear elsewhere; if I just include the theory for no reason whatsoever in my text, I am failing my reader - he or she has no chance of figuring out what significance I ascribe to the things I mention.

This also relates to clarity - the reader should be able to understand the meaning of the text without having to play a game with the author or performing feats of telepathy. Constantly having to second-guess without any way of verifying when the second-guessing was right or wrong is of no use.

If we assumed this essay already had reached its end, there would be an observation about it to add: I violate the maxim of quantity and maybe that of relevance; there is no reason to bring up neural networks at the point I brought them up. They are there as an empty mumbo-jumbo statement that in no way contributes to what I am saying; I invoke them as though just invoking them made my argument right - as if it were some kind of magical invocation. There probably would exist any number of better psychological things to look at to realize why we occasionally misinterpret what we read or hear. However, I have no background whatsoever in psychology, although I have taken classes on neural networks from a computer science point of view. In retrospect, it does serve a role - it illustrates how an irrelevant point can seem relevant to an unwary reader.

In a follow-up post, probably tomorrow, I will go on to explain how this makes my reading of Acharya's books reasonable even when her fans accuse me of maliciously misrepresenting her. Their explanations of the claims either rely on information not present in the books, on cutting out entire bits of sentences, and on peculiar reasoning that operate on the assumption that Acharya expresses her points in extremely unclear wordings.

No comments:

Post a Comment