«   |   »

Mind, syntax, semantics

by Neil Rickert

While thinking about the implications of a recent post it occurred to me that philosophy is almost an entirely syntactic enterprise, and pays little more than lip service to semantics.  To me, this was a surprising realization.  No doubt it explains why my own ideas are very different from those expressed by philosophers.  For I have long considered semantics to be the primary concern.

For perspective, let’s start with the Shannon-Weaver model of communication:

  1. We start with something meaningful, which we describe as semantic.
  2. The meanings or semantics are encoded into a stream of symbols, often taken to be binary digits.  This is the encoding step.
  3. The stream of symbols is what Shannon referred to as “information”, and is often called “Shannon information”.  Shannon’s information theory deals with the communication and manipulation of the symbol stream, and whatever meaning was encoded in that symbol stream is not relevant to the use of the Shannon information.  The symbol stream is usually considered to be syntactic, with the symbols as syntactic elements.
  4. The final receiver(s) of the symbol stream decode it in order to recover the semantics.

When we consider natural language communication, the natural language statements fit the requirements for Shannon information.  For written communication, the alphabetic letters are the symbols, while for spoken communication the phonemes are the symbols.

Much of philosophy deals with representations, which are normally taken to be something like natural language statements and are variously referred to as propositions, statements, beliefs, mental attitudes.  These representations are said to be intentional, which implies that they are about something.  That is, the representations are taken to result from the encoding process in step 2 of the Shannon-Weaver model.  However, what is encoded appears to play no role.  Philosophers normally see representations as being used with logic.  But while the syntactic strings may encode semantics, the logic is done based on form so the encoded semantics is not actually relevant.  Fodor’s “Methodological solipsism” thesis helps to make this clear.

Truth is considered an important property of some representations, and one might wonder whether truth is an implicit reference to the semantics.  However, truth is usually taken to be “correspondence with the facts” where facts are in turn taken to be metaphysical entities that are in the form of representations (i.e. are syntactic entities).

Philosophers do, of course, discuss semantics.  However, they generally try to treat meanings as defined by the truth conditions of statements.  That amounts to attempting to provide a syntactic account of semantics.

When knowledge is defined as justified true belief, that is usually treated as if knowledge amounts to stored representations, which would be an entirely syntactic account of knowledge.  Perception is typically taken as a system that delivers mostly true representations.

The idea of AI (Artificial Intelligence) fits quite well here.  If everything cognitively important is syntactic, then it ought to be possible to automate that on a computer.

My own constrasting view

I see the mind as primarily semantic.  I see perception as providing us access to the semantic world.  This is more or less consistent with J.J. Gibson’s ideas on direct perception.  Thus perception gives us access to the semantics of the world, and we ourselves then select from that and encode our selection into syntactic form for communication with speech.  Of course the brain is used for this encoding, but it would be parts of the brain associated with language, rather than parts of the brain associated with perception.  I see knowledge as mostly in the form of abilities.  These would include the abilities to perceive those semantic aspects of the world that are of interest, as well as the ability to encode that into syntactic form to use in speech.

Related to this, I see truth as not being a criterion for perception because truth is a criterion for syntactic representations rather than for unrepresented semantics.  Similarly, I see no role for the storage or retrieval of representations and I am skeptical of the possibility of AI.

Posted on April 18, 2012 at 03:11 (UTC) in cognition, philosophy   |  RSS feed

Tags: , ,

7 Comments to “Mind, syntax, semantics”

    
  1. Lage
    April 20, 2012 at 13:37 (UTC)

    Neil,

    As I mentioned in your previous post titled: “Information Storage In the Brain”, there doesn’t appear to be any alternative explanation to adequately explain how learned experiences (memories with context) can exist without some form of information and/or representative storage. I do agree with you in part — that when it comes to our conscious brain and thus how we “think”, it is primarily a semantic engine, but it still uses syntax to store memories in the physical sense (otherwise meaningless neural network patterns on their own represent semantic information once they are placed in certain parts of the brain and connected to particular other networks). We can’t negate the syntactic attributes needed to assemble (or reassemble) the semantic information (memories) properly either. Perceptions on their own don’t have much context or semantic qualities until they are compared with the reference baseline of the brain (created by all the experiences thus far). Just as an infant experiencing certain perceptions, due to a lack of sufficient data points (minimal reference baseline), will not have much context to “know” what they are perceiving is supposed to mean (the semantic quality). Very early on when the conscious mind and self-awareness are still in developmental stages, it can’t have very significant meaning (other than “instinctual” meaning which isn’t based on experiences at all) — rather it appears that they are mostly experiencing perceptions in a certain order coded into some form of memory (syntactic in nature through the use of otherwise meaningless neural networks) rather than perceptions primarily having context or meaning (semantic in nature). Later on, when the brain develops, due to the way the brain can relate different experiences and information stored, now these previously meaningless perceptions start to have context and meaning (become more semantic in nature). For example, we could experience a series of perceptions presented to us out of their expected order (where syntactic attributes are mostly negated), if an experiment was designed fairly well to accomplish such, and we could make much more sense of it regardless of syntax than a developing infant brain could hope to do. That developing brain would not be up to the task, and would require much more syntactic fidelity (in terms of perception itself as well as it’s storage in the brain) until it has established a better background reference baseline to ascribe some meaning to those perceptions. Once that has been established, THEN that brain could rely on semantics more than syntax (in terms of thought processes). It did however require syntax first (early on in life) as well as for the information storage, and so we could say that we have semantic and syntactic “software” running at all times, where the semantic software becomes primary later in life. The syntax appears to me to be most important when the brain is first learning the pattern of temporal relationships between perceptions (very early in life) as this seems to be most important when ascribing meaning or semantic attributes to those perceptions (needed for the first unique experience or for several iterations until the semantic attribute is refined and remembered). If we designed AI in this manner where it uses syntax for storage AND recognizes patterns through repeated experience (to establish a semantic quality), then we’d have something that would more than likely work as the brain does. I feel the biggest problem is that AI engineers aren’t able to make self-learning software very easily (where the computer continues to re-program itself rather than just store new information) because the brain is so complex and takes us humans YEARS just to program (among other issues regarding parallel processing difficulties, etc.). We need to be able to have the computer reprogram itself (with syntactic methods) rather than just store new information, just as the brain creates new neural networks and promotes/inhibits the effectiveness of old ones. This would be analogous to a computer re-programming itself (deleting some code and/or adding new code) based on it’s initial program instructions. The only way I see feasible to do this is for the computer programmer to make software capable of recognizing new patterns and their consequences if not recognized through repetition and the equivalent of sensory instructions (if the AI unit is saying some phrase and the programmer slaps it on the hand OR says “wrong” then it has to reprogram itself or try a different algorithm), as well as have some baseline of information so it has a place to start (unless we are trying to mimic an infant brain with our AI, in which case less information storage is needed from the start and would be much more easy to accomplish). It seems that AI engineers aren’t trying to mimic an infant brain, they want to mimic a human brain that has had years of experience, reprogramming, fine-tuning, etc. Very difficult indeed. Regardless of semantic qualities present, in my opinion.

    Like

    • Neil Rickert
      April 20, 2012 at 18:51 (UTC)

      I do agree with you in part — that when it comes to our conscious brain and thus how we “think”, it is primarily a semantic engine, …

      At least there is some agreement.

      …, but it still uses syntax to store memories in the physical sense …

      And that’s where we disagree.

      I am not denying that there are physical changes to the brain. I am skeptical that those changes can be considered the storing of syntax. It seems to me that they are likely to be changes whose role is to modify our behavior, and I think that is better described as semantic.

      Just as an infant experiencing certain perceptions, due to a lack of sufficient data points (minimal reference baseline), will not have much context to “know” what they are perceiving is supposed to mean (the semantic quality).

      I don’t like that way of talking at all. The infant does not experience perceptions. The infant experiences its world and perceives its world.

      “The infant experiences perceptions” seems to suggest that some black box is mapping the world into syntactic elements and the infant experience that syntax. But I see the infant as directly dealing with the world, not with syntactic representations of the world. The child’s brain might be coming up with syntactic representations of part of the world. But I see “perceiving” as the action of coming up with those syntactic representations, rather than examining the resulting representations.

      Perceptions on their own don’t have much context or semantic qualities until they are compared with the reference baseline of the brain (created by all the experiences thus far).

      I see that as backwards. That is to say, we normally compare things semantically, rather than syntactically.

      Like

      • Lage
        April 27, 2012 at 14:28 (UTC)

        Neil,

        “I don’t like that way of talking at all. The infant does not experience perceptions. The infant experiences its world and perceives its world.”

        The common definition of perception is “the result or product of perceiving”.
        A common definition of experience is “the totality of the cognitions given by perception; all that is perceived, understood, and remembered.”

        So really, that is what we experience; perceptions (including time perception which is manufactured by the brain). The difference between the two is that our experience is thought to be the seemingly continuous totality of all the individual perceptions, but we are still experiencing perceptions. Different experiences are a result of perceiving things differently, thus having different perceptions. If you disagree, then state an example of something you have experienced, where that experience can’t be broken down into constituent perceptions. I think that will be difficult to do. Perhaps you are confusing the word “perception” with “sensation”, where you think that I mean that “an infant experiences sensations”, which is not nearly the same thing. However, the brain is thought to be the “black box” that translates the nervous system’s incoming “objective” sensory data (photons or matter interacting with the body in different ways) into subjective experience (our brain’s perception of those sensations).

        “But I see the infant as directly dealing with the world, not with syntactic representations of the world.”

        Are you trying to suggest that our experiences are objective in some way? We have no right to assume that our experiences are objective in any way, so to say that the infant is “directly dealing with the world” requires more clarification. For all practical purposes, we can consider “the infant” to be the infant’s brain (or “mind” to be more specific). The brain of that infant does represent the world by taking syntactic streams of data (sensory data incoming in a particular order) and converting it into a subjective experience that is beneficial for it’s survival (overall through Natural Selection). The consensus in science is that objective data (like photons, matter, etc.) interact with our body via the sensory system (nervous system, etc.) converting a truncated amount of incoming data (due to limitations on nervous system processing speed and resolution) into even further truncated streams of information (due to nervous system compression before & as a result of space limitations in the body/spinal cord), eventually leading to the brain where that truncated data is translated into what we perceive (perceptions). Our perception of the world is completely different than that of another animal with a completely different brain, as it is that very brain that produces the perceptions. This implies that every brain “deals” with a different representation of the world even if the sensations (incoming sensory data) were equivalent — but each brain is initially dealing with (translating) a syntactic representation of the world (incoming sensory data). What we experience is semantic in nature, but it results from the brain converting that syntactic information (sensation) into semantic information (perception/experience).

        Like

        • Neil Rickert
          April 27, 2012 at 17:44 (UTC)

          The common definition of perception is “the result or product of perceiving”.

          Of course, I checked in some dictionaries, and that is not what I found. What you give is perhaps how “percept” is defined. However, I see “percept” as a completely unnecessary word, formed via an inappropriate reification.

          A common definition of experience is “the totality of the cognitions given by perception; all that is perceived, understood, and remembered.”

          I didn’t find anything close to that. It seems obviously wrong.

          Are you trying to suggest that our experiences are objective in some way?

          I’m not even sure what is being asked there. I wonder what you mean by “objective”.

          I see experience as being inherently subjective. And I see the objective as derived from the subjective. Roughly speaking, we say that something is objective when it agrees with the consensus view and does not depend on individual biases.

          As for the rest of your comment – if people really think that way, then it is no wonder that there is so little progress in studying cognition. It seems to be a thoroughly dualist theory of mind, from which the spiritual soul has been thrown out. But all that remains is a theory if meat. There’s nothing left of cognition, except a few words used as fillers. You need to toss that into the trash can, and start over afresh without all of those old dualist assumptions.

          I am thinking about doing a full post on that. It wouldn’t be trying to pick your statements apart. Rather, it would be attempting to present an entirely different way of looking at things. (It might take more than one post).

          Like

          • Lage
            May 1, 2012 at 16:01 (UTC)

            “Of course, I checked in some dictionaries, and that is not what I found. What you give is perhaps how “percept” is defined. However, I see “percept” as a completely unnecessary word, formed via an inappropriate reification.”

            http://dictionary.reference.com/browse/perception
            The 3rd definition reads pretty clearly: “the result or product of perceiving”, so the dictionary will back me up on this one.

            As for my definition of “experience”, we have:
            “the totality of the cognitions given by perception; all that is perceived, understood, and remembered.”
            This is the 5th definition of “experience” given in the dictionary at:
            http://dictionary.reference.com/browse/experience?s=ts

            “I’m not even sure what is being asked there. I wonder what you mean by “objective”.

            When I mentioned the word “objective”, I was referring to true objectivity, that is — the way things really are regardless of our experience, as opposed to “subjective” (the way we think things are based on our unique experience). That was why I mentioned it. More specifically, it was to address your comment of the infant “directly dealing with the world”, which needed more clarification as one could infer that you are implying objectivity which I think is nonsense.

            “As for the rest of your comment – if people really think that way, then it is no wonder that there is so little progress in studying cognition. It seems to be a thoroughly dualist theory of mind, from which the spiritual soul has been thrown out. But all that remains is a theory if meat. There’s nothing left of cognition, except a few words used as fillers. You need to toss that into the trash can, and start over afresh without all of those old dualist assumptions.”

            I’m interested in knowing what specific part of “the rest of my comment” you are referring to?

            -Peace and Love

            Like

          • Neil Rickert
            May 1, 2012 at 17:59 (UTC)

            http://dictionary.reference.com/browse/perception

            That is one of the dictionaries that I checked. Most dictionaries give their definitions in order of importance (in terms of usage). That you had to go to definition 3 in one case and definition 5 in another, already shows you are depending on unconventional usage.

            When I mentioned the word “objective”, I was referring to true objectivity, that is — the way things really are regardless of our experience, as opposed to “subjective” (the way we think things are based on our unique experience).

            I can see that I need to do a post on “objective.” Nothing that we know is “regardless of experience”, so I see objective as arising out of the subjective.

            I’m interested in knowing what specific part of “the rest of my comment” you are referring to?

            I have now made a separate post based on that “rest of the comment”.

            Like

  2. Lage
    May 1, 2012 at 18:38 (UTC)

    “That you had to go to definition 3 in one case and definition 5 in another, already shows you are depending on unconventional usage.”

    It does not matter if it was the first or the tenth definition in the dictionary. If it’s there, it counts as an agreed upon definition (even if used less often). My definitions chosen are also very intuitive.

    “I can see that I need to do a post on “objective.” Nothing that we know is “regardless of experience”, so I see objective as arising out of the subjective.”

    I agree with you that nothing we “know” is “regardless of experience”, and I never argued otherwise. I was referring to true objectivity, that is, the way things really are outside of our subjective experience. How the world really is (i.e. What’s really “out there”), not just how we humans perceive it (let alone the different experiences we have from person to person). We are discussing two different levels of objectivity which explains why you were confused. I agree that what many think they know to be “objective” is a result of arising out of the subjective. The problem is that we can’t actually know anything objectively about the way things are outside our perspective because all we have to analyze the world with is a subjective experience. We may say that if multiple people agree that something is an objective fact, then you have “objectivity”, but in reality, if all those perspectives are subjective, then just because we have an agreement on our subjective experiences doesn’t mean that what we think we know IS an objective truth about reality. It’s a poor convention in how we use the term “objective”. That’s OK though.

    Like

    
%d bloggers like this: