Monday, 19 June 2017

Language: Two Himalayan Mistakes

Seated Woman by Richard Diebenkorn
Posted by Thomas Scarborough
We take a lot on trust. Too much of it, mistakenly. We even have a name for it: ex verecundiam.  With this in mind, there are two things at the heart of our language, which we have mistakenly taken on trust. The first is how to circumscribe the meaning of a word, the second is how to qualify that meaning. These are not merely issues of semantics. They have profound implications for our understanding of the world. 
There was a time, not too long ago, when we had no dictionaries. In fact, it was not too long ago that we had no printing presses on which to print them. Then, when dictionaries arrived, we decided that words had definitions, and that, where applicable, each of these definitions held the fewest possible semantic features. A woman, for instance, was an ‘adult human female’, no less, and certainly no more – three features in all. While this may be too simple a description of the matter, the meaning will be clear.

We may never know who first gave us permission to do this, or on whose authority it was decided. It may go back to Aristotle. But at some time in our history, two options lay before us. One was to reduce the meaning of a word to the fewest possible semantic features. The other was to include in it every possible semantic feature. We know now what the decision was. We chose artificially and arbitrarily to radically reduce what words are.

We canvassed the literature. We canvassed the people. All had their own vast ideas and experiences about a word. Then we sought the word's pure essence, its abstract core – like the definition of the woman, an ‘adult human female’. This, however, introduced one of the biggest problems of semantics. We needed now to separate semantic features which mattered from those which did not. The artificiality and uncertainty of this dividing line – that is, between denotation and connotation – has filled many books.

Worse than this. It is easy to prove that we took the wrong option at the start.  We are in a position to demonstrate that, when we refer to a word, we refer to its maximal semantic content, not minimal. Some simple experiments prove the point. Take the sentences, ‘I entered the house. The karma was bad,’ or, ‘The car hit a ditch. The axle broke.’ What now does ‘the karma’ or ‘the axle’ refer to? It refers to the maximal content of a word. This is how, intuitively, innately, we deal with words.

Our second big mistake, which follows on from the first, was the notion of subject and predicate. We call these the ‘principal syntactic elements’ of language. They were at the forefront of Kant's philosophy. Today, the universally accepted view is that the predicate completes an idea about the subject. Take as an example the sentence, ‘’The woman (subject) dances (predicate),’ or, ‘The penny (subject) drops (predicate).’

Again, ‘the woman’ is taken as the bare-bones concept, ‘adult human female’. Add to this the predicate – the fact that the woman dances – and we expand on the concept of a woman. We already know that a woman dances, of course. We know, too, that she laughs, sleeps, eats, and a great deal more. Similarly, we define ‘the penny’ as a ‘British bronze coin’. Add to this that it can drop, and we have expanded on the concept of a penny. Of course, we know well that it clinks, shines, even melts, and much more besides.

Yet, what if the predicate serves not to expand upon the subject, but to narrow it down? In fact, if words contain every possible semantic feature, so too must subjects. A predicate takes a ‘maximal’ subject, then – the near infinite possibilities contained in ‘the woman’, or ‘the penny’ – and channels them, so to speak. ‘The woman (who can be anything) dances.’ ‘The penny (which offers a multitude of possibilities) drops.’  Predicates, then, are ‘clarifiers’, as it were. They take a thing, and narrow it down and sharpen its contours.

The application to philosophy is simple.  We discard a word’s many possibilities – those of a woman, a penny, a house, a car – in the interests of the arbitrary notion that they represent minimal meanings – so reducing them to the smallest number of semantic features people use, and throwing the rest away.

Day after day, we do this, through force of centuries of habit. With this, we instantly discard (almost) all the possibilities of a word. We meet situations without being open to their possibilities, but cobble a few predicates to bare-bones subjects, and so lose our good sense. Nuclear power is the generation of electricity, a ship is something that floats, a F├╝hrer is someone who governs. The words, being stripped of their maximal meanings, do not contain – perhaps most importantly – the possibility of evil. This greatly assists prejudice, bigotry, partiality, and discrimination.

When words are reduced to their minimal features – when we base their meaning on their denotative core – we ‘crop’ them, truncate them, reduce them, and above all, cut away from them a great many meanings which they hold, and so reduce our awareness of the world, and cosmos.  Due in no small part to the way we imagine our language to be – minimal words and minimal subjects – we have entered habits of thinking which are simplistic, reductionistic, technical – and dangerous.

But to understand words in terms of maximal meanings is to reject the reductionism of our present time, and to think expansively, creatively, intuitively, holistically. 


  1. The essay, an important look at language, ends with this interesting teaser: ‘When words are reduced . . . to their denotative core . . . we reduce our awareness of an inter-connected world, and cosmos’. Might you expand on this idea a mite? Specifically, how do you see denotation reducing our awareness of an ‘inter-connected world, and cosmos’? And ipso facto, how does connotation presumably increase our awareness of inter-connectedness?

  2. A teaser typically refers to something which is yet to be said. My closing paragraphs refer to what has already been said. At least, they are intended to. They represent a conclusion. I base the conclusion on two principal observations:

    • our exclusion of semantic features from our vocabulary, and
    • our stripping of semantic features before we use a predicate

    While I dislike empirical debate, because of its potentially vast complexity, a lot of research has been done which links good judgement with ‘flexibly bounded categories’ and the like. One example, Gislason.

    Whorf is remembered for his ‘denotative’ petrol drums. Yet he gave many more such examples in ‘The Relation of Habitual Thought and Behaviour to Language’. Whorf, however, is unsafe waters! Merely for amusement.

    I’ll gladly give the question another attempt if the reply seems to be beside the point.

    1. As your last query itself hints at, yes, your ‘reply [does] seem to be beside the point’. What I was getting at is that I don’t see how the post supports the leap from the mixed assortment of fairly narrow topics discussed in the post’s body — like subjects, predicates, semantic content, truncation, and dictionaries — to the orders-of-magnitude-more-bold conclusion of ‘reduc[ing] our awareness of the world, and cosmos’. I suggest the discussion doesn’t get there from here.

      But let’s move on. I’ll narrow further comment to just one aspect of language that the essay explores to some degree, which is ‘denotation’ and ‘connotation’.

      At the peril of being obvious, both denotation — the ‘smallest number of semantic features’ — and connotation — variegated layers of rich meaning — have an equally rightful place. That is, if used in the proper context — apart from the presence or absence of ‘predicates’. Surely denotation, in ‘circumscribing’ and ‘truncating’ the meaning of words, is indispensable — in denotation’s rigor, lessening the chance of ambiguity and misunderstanding.

      By way of a simple example, if a drooling patron orders a fresh ‘pain au chocolat’, ‘circumscription’ and ‘truncation’ (to borrow terms from the essay) make sure she won’t be handed a glistening mound of ‘steak tartare’ instead. (Did I mention she’s vegetarian?) More consequentially, when two plasma physicists discuss — with deliberately technical circumscription and precision — the use of deuterium and tritium in the generation of fusion power, their resulting unmitigated clarity is a good thing. As, much less direly, when I want to buy a half-kilo of parsnips or a bunch of festive party balloons.

      In these and endless other everyday circumstances, dictionaries, in prescribing meaning and usage, serve as an essential tool, to help avoid chaos in language. Dictionaries, through their definitions, help frame some of the shared conventions of language — as opposed to meanings that, left to individuals to choose, are therefore random and haphazard. (An aside: I do find Samuel Johnson’s dictionary definition of ‘lexicographer’ as ‘harmless drudge’ a delicious instance of self-deprecation.)

      Anyway, I venture to say that one can still think ‘expansively, creatively, intuitively, holistically’, even when circumstances call for minimising use of connotation and maximising resort to denotation — including in philosophy, Pi essays to boot. Denotation, with its ‘semantic minimal content’, does not so much entail ‘throwing away’ those important variegated layers of meaning, but merely temporarily shelving them. To be called upon later — in the right circumstance, for the right purpose, with nuance and subtlety in mind, and still with clarity.

    2. I think the question is perceptive, if I may call it one question. I have thought a lot about it, and have written a lot about it (unpublished).

      What you write in conclusion is correct – only the way that we go about with language, we tend not to ‘temporarily shelve’ connotations*, but permanently shelve them. This is the problem with (physical) science, which necessarily ‘screens things out’ in order to progress (and must use strictly normed language to do so), yet fails in most circumstances to consider its re-entry into the world – however some greater care is taken in cases such as pharmaceuticals or genetic engineering. Generally speaking, science stays very much on the level of denotation, so to speak.

      I have likened it to catching a curious insect in a bottle. After studying its behaviour inside the bottle for a few days (with 'denotational', scientific language), and faithfullly recording everything that it does there, we now release it into a new environment. Of course, we do not know quite what it will do there. It might die and return to the ground, or we might visit the same environment in fifty years’ time, to find that it has completely changed. But I am thinking not only of environments. I am intending, too, systems such as economics and society and so on.

      Or one could put it all quite differently: we are discussing here what definitions are. The received account would hold that definitions are the exact meanings of words, and these exact meanings equal a word’s minimal features – perhaps, too, the relationships between these features. This is too simple, but it is basically true. But if one does away with the distinction between denotation and connotation, then one must say that definitions are something like this: they describe relational complexes within a word – or rather, priority relational complexes**. But one cannot speak of denotation.

      As things are, we live in a great dichotomy about language, which I think your question reveals. We use words one way, yet we define them another way. But we accept the tension. Yes, a word is its maximal meaning. No, a word is its minimal meaning. No wait, it is both. But that can’t be. But where do we now draw the line, if we accept that denotation exists? I would hold that the dichotomy between denotation and connotation can be done away with, and a lot of problems would be solved at the same time.

      It is, however, a big subject. I have ‘given it a go’ here briefly, and I hope I have made sense.


      * I use the word ‘connotations’, although I do not think a distinction exists between denotations and connotations.

      **Einstein called them complexes from nature.

  3. Dear Thomas,

    I wonder about the willingness to strip away words of their habitual meaning.Often when this is explored the reader says: 'this is vague' and as consequence 'the subject is not clear' which involves an acceptance of non comprehending a context in how we are educated to comprehend it. The problem hides in the expectation of the words. As long as this projection cannot be 'uneducated', to think expansively, creatively, intuitively, holistically, will be rather difficult. The lack is not within the signs, but in us. Would a revision of semantic features offer a different way of thinking? Or maybe the thinking is not just about words?

    1. Yes, I think this is perceptive. But what does it mean? All over, one reads of a decline in critical thinking and creativity. Language, which has lost its 'free flight', is often said to play a big part in that.

      Assuming that words do represent 'maximal content', then every word we speak potentially conjures up the whole universe as it were. I do think we can reduce or increase our awareness of 'the universe' as we speak, and being aware of how we misinterpret language could be a start.

      Do we think without words? That is a classic question. I don't remember whether anyone answered it yet.

  4. It is 'ironic' that a piece in which you protest against dictionaries for hiding information - specifically the full range of senses and meanings of a particular word, privileging a few, you make repeated use of obscure terms and references.

    "Yet, what if the predicate serves not to expand upon the subject, but to narrow it down? In fact, if words contain every possible semantic feature, so too must subjects."


    And what I greatly admire about Whorf, who you dismiss too easily as 'unsafe' is the elegance and clarity of his arguments. The oil drum one you mention, again without eleborating!, seems to me a very reliable one. Whorf says the use of the word 'empty' on a sign by metal gasoline drums led workers to think of the drums as 'empty', void, 'full of nothing', when in reality the drums *were* full - of inflammable residues.

    Woould Thomas have dictionaries offer 'full' as an extra synonym for 'empty, though?!

    But better than answer that, yes, it would be very interesting to have Thomas look at that last question here sometime: 'Do we think without words?'

  5. Back in the day, Martin, subject and predicate were not THAT obscure.

    Whorf has his merits, and personally I believe him, but he seems like sinking sands for a writer. For further examples Whorf offered, apart from the petrol drums, readers may see 'The Name of the Situation as Affecting Behavior' at

    OED: empty = containing nothing
    SED (Scarborough English dictionary): empty = purportedly or visibly containing nothing, however in likelihood containing a great deal, possibly of fatal significance. For example of such a thing, see 'Language: Two Himalayan Mistakes'.

  6. I came across an interesting endnote today in Francis Schaeffer, which has a bearing on this post: 'My thinking has led me to believe that there is a collective cultural consciousness or memory which is related to words. ... This is the case regardless of the individual's personal world-view and despite what the dictionary of scientific textbook definition have become.'