Saturday, October 12, 2013

Concepts

On 2013-10-12 9:02 AM, Matthias Melcher wrote:
If we consider, on one hand, all the aggregated connection
patterns that make up a simple concept or word in an individual,
and OTOH, in a society or a community of practice, we consider
the small part of knowledge of that society that roughly
corresponds to one term used by this community, can we say that
these two are very similar, except for the notorious 3rd person
vs. 1st person distinction?

(If the question is not relevant for you, please feel free to just
tell me, because even this would already be a valuable hint.)
We need to be more precise about what you mean when you say "can we say that these two are very similar".

If you mean "the concept of 'Paris' in an individual is similar to the concept of 'Paris' in a society", then although there is an overlapping association with the word and sound 'Paris' I have no reason to think that a person's concept and society's concept is similar at all. But even here, it depends on how we define 'similar'... my own thinking of 'similarity' is roughly the same as Tversky's feature-based analysis of similarity, however I additionally incorporate a concept of salience, following Stalnaker's employment of the term in modal logic. But what is a 'feature', in this analysis? In a human, a feature is an aspect of sensory awareness, what the positivists would have called a 'sense datum' (but without the presumption of logical atomism; it's just a sensation, a percept, not an atomic proposition). What would society's analogous 'sense datum' be? How does a society perceive? I would be interested in an answer to that question.

If you mean "the notion of a 'concept' for an individual is similar to the notion of a 'concept' for society" then I would agree. A concept in an individual - loosely speaking - is a subset of the full set of connections in the human brain (and to 'think' of the concept is to have at a given time that subset, or a sufficient part of that subset, electrochemically activated (ie., 'spiking'). A concept in a society is the same sort of thing - depending on how exactly we want to construe it, it is the connection of a subset of the set of all individuals in society (*), and society 'thinking' of the concept is the activation of that subset, or a sufficient part of that subset. For example, a concept similar to 'rage against the machine' might be active when enough people open their windows and shout "I'm mad as hell and I'm not going to take it anymore." The spreading of a meme (see my paper 'Hacking Memes') through speech, print, artifacts, text, etc., is an instance of society 'thinking' of a concept. Though tat said, probably social concepts are far more complex than that - it is very unlikely the concepts a society 'thinks' are very similar to the concepts an individual thinks (as I suggested in the first paragraph).

(*) We can have different accounts of 'social concepts' depending on how we define the concept of a social 'node'. If we define it as individuals only, then a concept requires the activation of individuals, and would be expressed eg. by what individuals say, do, etc. - a 'culture' on this account is essentially a set of social concepts. If we expand our set of individuals to include artifacts, then we can have a much wider sense of social concepts, particularly when the artifacts are not simply static, like books and paintings, but dynamic and interacting, like stock futures, financial exchanges, highways systems, etc. We can extend the idea of social concepts by a third degree is we allow the set of individuals to include non-physical entities, such as words and concepts; now the set of interconnected entities includes the associations of words and other non-physical entities with each other (though this would have to happen in a physical substrate).

If this isn't clear to this point, stop here, because the next bit will really mess you up.

It's this: no set of connections, or activation of a set of connections, is inherently a 'concept'. What we call a concept (and similarly, what we call an 'idea') does not exist inherently in nature, but exists only insofar as it is perceived and recognized as such. For example: a set of points on the surface of Mars might be connected and inter-related, and when the light shines on them, activated (by reflecting light), and hence form the physical conditions necessary to instantiate a concept. But it takes a third party - specifically, another set of connected entities (a neural network, in this case) - to recognize that this set of points on Mars looks like Jesus, and hence instantiates the concept 'face of Jesus'. An entity (or set of connected entities) X is 'recognized' as the concept P when perception of X causes a sufficient set of activations of a network to cause the full set of activations of the connections constituting P to occur in the neural network. Or to make the same point more colloquially: a bunch of mountains on Mars is called 'the Face of Jesus' when a person looking at the light reflecting off the mountains sees enough of the face to think "hey, that's the face of Jesus."

What we call consciousness or sentience is the capacity to do this in ourselves, that it, the activation of a set of connections in one part of our brain is sufficient to cause the recognition process to happen in another part of our brain. So we have the perception of some orange in a background of green, and this causes the full set of connections constituting the concept 'tiger' to be activated more deeply in our brain. (It can also work in reverse: when the connections constituting the concept 'tiger' are activated, for whatever reason, we can sometimes have the experience of perceiving orange and black stripes).

So why am I hesitant to use the word 'concept'? It's simply because the word 'concept' has so much baggage attached to it. People see the word 'concept' and they start thinking of intentionality and meaning, things out there in the world, sets o logical propositions, models, and more. There is an entire set of (fictional) artifacts built around the concept of 'concept', and I don't mean any of them when I use the term.

4 comments:

  1. reminds me of that quote from Wittgenstein: 'concept' is a vague concept !

    ReplyDelete
  2. Thanks very much for your detailed disagreement. Indeed I meant the former of your two alternatives. Maybe my thinking was different because I did not see the society as an entity that may have "1st person" experiences such as moods and rages -- in this case it is clear that e.g. for the society of France, "Paris" is "my heart" and not just another large city as for an arbitrary individual.

    When I thought of a society or a community of practice as a whole entity, I thought of it not as a 1st person but rather as a 2nd person, as one to interact with and exchange with. And in such exchanges, concepts of both parties must be at least so similar that communication does not always fail.

    If society is a very abstract 1st person, it is also very difficult to guess how it "perceives". But without perception, no recognition would be possible and hence no knowledge, right? So I'll try to guess, and I would say that the community perceives, most importantly, all the utterances of its members, and each of these contributes a data dot in a large picture on the "retina" of the community.

    Even though this sort of perception is probably not optimized for distinct "features", it allows for proximity and for weak and strong connections. And maybe it could also provide the missing physical substrate for "the associations of words and other non-physical entities with each other"?

    ReplyDelete
  3. A concept is the smallest unit of cogntive meaning. Similar to a phoneme in language, a concept represents a single cognitive space. Essential to this definition of concept is the understanding of a unit of meaning. Regardless of the complexity, a unit of meaning must stand alone as thought. It may require additional concepts to trigger action or create connection, but all concepts exist as a result of an individual's unique construction.

    ReplyDelete
  4. @Jenny, I am not sure about "stand alone as thought", because I never experience thoughts as "alone" in the sense of "disconnected". Perhaps "singled out" by distinguishing one concept from all others? Indeed it may be the smallest "unit" that is distinguishable from others. Similarly like two shades of blueish green are not distinguishable from one another by the human perception while two others are?

    ReplyDelete

Your comments will be moderated. Sorry, but it's not a nice world out there.