Wednesday 13 May 2015

Language In A Petri Dish: the scientific misunderstanding of signals



No sane person deliberately seeks to misinterpret messages or to incorporate falsehoods into their reasoning. We pay attention to symbols because the consequences of ignorance or misunderstanding can be disastrous. Driving through a red traffic signal is an act that thankfully almost all drivers wisely avoid. If we weren't careful about the ways that we use symbols — their meanings — then communication would quickly descend into incomprehensible babble. It really does matter how we use signs and for the most part we stick to the rules. But sometimes even scientists are sloppy. This post is about a very specific but widespread form of scientific sloppiness: the misattribution of symbol-use to cells and simple organisms. 

Symbol users act in extremely strange ways. On the basis of a simple sign — a word, a coloured light or an abstract scrawl — we can be led to engage in some of the most elaborate, sophisticated, and sometimes the most bizarre, behaviours. And perhaps the most bizarre thing of all, is that the sign itself can be formed from absolutely anything. That is the extraordinary power of symbols: we can use anything to symbolise anything else, so long as the people we are communicating with know the rule we are using.

Rule following is perhaps the most fundamental requirement of symbol use. If we do not know the rule, we cannot know how to respond. This is why only the most intelligent creatures are capable of using symbols — because only the most intelligent creatures are capable of using tools; of putting raw materials to uses for which they were never designed.

A very brief scan of current research within the biological sciences will be sufficient to demonstrate that talk of chemical "signalling" between organisms (and even between cells) is extremely common. And neuroscience is almost entirely committed to the conviction that neurons produce signals. In my experience the merest suggestion that such talk is mistaken is often regarded as tantamount to heresy, not because there are particularly compelling reasons for supporting such a view, but because there seem to be so few reasons against it. In other words, talk of biological signalling is simply a terminological habit or convenience that adds just a little glamour to terms that would otherwise be restricted to "stimuli" and "causal triggers".

I am sometimes asked whether I think it does any harm to talk of biological or neurological signalling. My usual response is to say that I’m not in a position to know. But a better response would be this: what good does it do to suggest that we can observe the rudiments of language in a petri dish? Since when was it wise for scientists to get it manifestly wrong about their philosophical foundations, and since when was it wise for philosophy to follow suit?

277 comments:

«Oldest   ‹Older   1 – 200 of 277   Newer›   Newest»
Andre said...

Jim, we really need to talk about some of the assumptions that underly your own thinking here. Like "because only the most intelligent creatures are capable of using tools; of putting raw materials to uses for which they were never designed."

When I read your posts I often get the feeling that you haven't really questioned your own assumptions underlying your position. I could be wrong, of course ..... But maybe a discussion would show which is which ......

Jim Hamlyn said...

You speak as though you have found more than one weakness in my view, Andre. I agree that the sentence you quote is the most obviously question begging one in the piece but can you substantiate your opposition or critique more fully please, because I'm not at all certain that I know what grounds you have to suppose that cells and microorganisms trade in symbols or what other apparent weaknesses you observe in my position?

Am I to assume that you think scientists are more justified in their foundational assumptions about symbol use than I am because (as far as you can see) I have not "really questioned my assumptions" whereas (according to you) they have?

As to the question of tool use. Show me a creature that is capable of using symbols that isn't capable of manipulating raw materials or conspecifics to achieve a goal.

Jim Hamlyn said...

In fact, show me a creature that we do not suppose can form a representation of a goal that isn't also what we would thus define as an intelligent creature?

Andre said...

What do you see as a signal?

Jan said...

What does this even mean? What are biologists wrong about? Why does it matter?

Jim Hamlyn said...

Did you actually read it Jan? What is so hard to work out about what it means? It means that biologists are working under a mistaken assumption that a chemical trigger or electrical impulse constitutes a signal. So you don't think it matters that biologists get it right about what cells and microorganisms do?

Andre, a signal is a form of communication in which one thing (a symbol) is used to represent another.

Jan said...

Jim - yes, every last word. I thought it was an introduction to a larger article and was very surprised when I realized that was it.

You basically just said 'they are philosophically mistaken' without explaining why, what that means, or why anyone should care. You indirectly gave some vague definition of a 'symbol' (which was quite lacking, a basic course in lexicology would provide much more insight into concepts such as context and creation of symbols), implied that biologists use the same definition, and simply assumed they were wrong in doing so.

This article is basically the best possible demonstration of why philosophy has negative connotations in the scientific community.

Jim Hamlyn said...

There is plenty more on the blog for anyone who wants to look further into the issues. The basic premise is simple yes, but that doesn't mean that it is mistaken. Your comments do nothing to refute the central claim. If you think it takes an understanding of lexicology to know what cells do or do not do then this must apply to biologists also. Lexicology is irrelevant to the issue of whether microorganisms and cells use symbols.

So what definition are biologists using Jan? How, and moreover what, explanatory purpose does it serve to invent (if that indeed is what is happening, which I would refute) a nonstandard usage for the words "signal", "sign", "message", "communicate" etc? Are you trying to tell me that philosophers who also use these terms are also using them in a technical sense to mean "causal trigger" and "impulse" as opposed to the more complex meanings implied? Again, what possible explanatory purpose does such usage confer?

Thomas said...

I think Jan quite clearly provided a refutation of your central claim - "You basically just said 'they are philosophically mistaken' without explaining why, what that means, or why anyone should care."
He pointed out you had only vaguely defined 'signal' as you intended it (and pointed you to lexicology for a broader definition of symbol in that sort of sense, in no way implying lexicology had any relation to the biologists use of 'signal' as you seemed to assume). Then noted you hadn't given any evidence of how biologists intend signal, just assumed they intended the same meaning. Then just assumed that would lead them to errors with no explanation of why let alone any evidence of failures because of this (IYO) fundamental terminological misuse.

"How, and moreover what, explanatory purpose does it serve to invent (if that indeed is what is happening, which I would refute) a nonstandard usage for the words "signal", "sign", "message", "communicate" etc?"
I'm sure you are aware Jim that many technical fields use terms in non-standard ways, using standard words for technical concepts. Or are you next going to attack mathematicians for their clearly misguided use of the term 'plane', how many students must keep wondering where the wings are? More seriously, electrical engineers talk of signal processing where signal just means an electrical waveform (in fact I think this is closely related to the definition in biology though they also include chemical signals). Are they also misguided?

As to "what definition are biologists using": I'd say they are using an operational one. Whatever experimental techniques they use to verify signalling are what signalling means to them. If the use of the term signal with its other connotations in general usage led to bad operationalising then you might have a decent argument. I see no evidence from you there. Similarly if you had evidence that the talk of signalling led to them imputing other characteristics to cells that didn't pertain *and this led to further experimental failings* then again you might have a point. Again I see no evidence from you there.

By failing to engage with what biologists actual mean by signalling, and failing to show any evidence of failings within biology, I think you reduce your argument to mere linguistic policing, at best "lexicographical purism" (to borrow from your quote of Dennett). I agree with Jan that philosophers sitting in their arm chair trying to police the language of scientists with no understanding of how the language is actually used in the lab is "why philosophy has negative connotations in the scientific community".

Jim Hamlyn said...

Ok well let me put the question differently: what scientific benefit arises from flouting lexicographical purism?

Jan said...

The task of the lexic in scientific communities is enabling precise communication. Your article is essentially just a misunderstanding of the role that context plays in this process - i.e. it is not a philosophical problem for scientists, but your personal problem of not understanding some linguistic concepts.

We can of course discuss the problem of compactness vs. precision of the lexic - but that's a linguistic discussion, which is not what you started here.

You just have to accept that words have multiple meanings, which are context-dependent.

Thomas said...

Jim: I'd say the function is communicative expediency. It means you don't have to keep making up completely new words for technical concepts. Plus by using existing words whose common meanings are sufficiently close to the technical usage it is fairly easy for a reader new to the area to get what is going on.
'OK, this cellular signalling it's going to be something like normal signalling, carrying information or something similar' versus 'What the fuck is cellular gloquaxing'.

Jan said...

Thomas yes, I agree, look at my second paragraph. I just think that's a separate discussion.

Thomas said...

Jan: Sorry, that was a reply to Jim, not you. Totally agree that is somewhat separate to the original discussion. Though this issue keeps coming up in discussions with Jim who makes a lot of similar attacks on scientific usage of terms that he doesn't agree with (all around the issue of representation/meaning/signalling, he isn't *just* a lexicographical purist).

Jim Hamlyn said...

When I dip blue litmus paper in an acid it turns red. I take this to be a sign (or a signal, or an indication, or a message, or an intimation, or an indication, or .......whatever) that what I have just dipped my litmus paper into is acidic.

That's OK in anybody's language, whether a scientist or a layperson. But who wants to say that this liquid is sending me a signal that it is acidic? (as a sympathetic warning to me, perhaps, not to drink it?)

Maybe you want to say that. Maybe some scientist wants to say that. But I don't want to say that. I want to say that signalling is a purposeful action and flasks of liquid are incapable of performing purposeful actions.

Thomas said...

"That's OK in anybody's language, whether a scientist or a layperson."
No, it's not. The litmus paper isn't a signal in the electrical engineering sense. The electrical engineer (if he was silly enough to think your usage of the term signal was the same as his) could claim you were completely wrong there. As could the biologist. Your usage of signal doesn't accord with theirs. Of course I'm sure most biologists and electrical engineers would quickly realise from context that you meant a different sense of signal. Why can't you see the converse?

Jan said...

The difference is that the brain and cells embed information-processing systems within them, and thus we can speak of signals as transfers of information.

When cells receive a certain protein on their membrane and change their behaviour according to this information - we can safely say that other cells have 'signaled' something.

I disagree that a connotation of purposefulness is a necessary semantic component of the word 'signaling' - it can be used in the context of any sufficiently complex and interacting system (like a computer program, a brain, or other tissue).

Matt said...

I think the word "sparticle" is a prime example of why you should avoid letting scientists come up with their own words for scientific stuff.

Also, scientists tend to know what they're talking about when communicating in practice. Just because some philosophers and lay people get confused, doesn't mean they need to be taken away by the language police.

Jim Hamlyn said...

I'm not suggesting that scientists need to invent new terms. But nor am I convinced that "communicative expediency" constitutes justified grounds for accepting usages that come laced with further unjustified implications — especially where more passive terminology and locutions are available. Accepting this kind of usage at the foundational level simply underwrites the illegitimate assumption of neural representation. And who here wants to tell me that alleged neural representation is not really representation but just a nonstandard lexicographical usage that has nothing to do with any other kind of representation we know of?

Jim Hamlyn said...

Thomas,

Do you believe that purposeful actions are performable?

If not, then what's wrong with ordinary language? Is it not a purposeful action when you signal to the waiter by lifting up your empty wine bottle? (The argument that what seem to be purposeful actions are, in reality, just causally determinate behaviours is, of course, profoundly philosophical).

If purposeful actions (a tautology, incidentally) are possible, then what sorts of things are capable performing such actions? Perceiving organisms such as ourselves? any non-perceiving organisms a bit smarter than a turnip? Any non-perceiving organisms whatsoever, no matter how rudimentary? Any mechanisms at all? Any lump of rock?

Don't scientists need to think about this before they speak with radical carelesslness about signalling in relation to mechanisms or to non-perceiving organisms or to perceiving organisms; and then defend their conceptual confusion by protesting that they are as entitled as anybody else to modify the use of the verb 'to signal' in any way they please, whether or not it commits them (unthinkingly) to an obliteration of the distinction between regularly causally explicable behaviours and purposeful actions?

Jim Hamlyn said...

If scientists only "tend to know what they're talking about" then it must be the case that there are times when they do not know what they are talking about or, worse still, they assume that they know what they are talking about but in fact they do not. What good does it do to use misleading expressions in scientific research and what are the repercussions in other fields of enquiry that refer to scientific studies in which the careless use of language is commonplace?

Why would anyone who is careful in every other respect of their enquiries wish to treat the description of their findings with such negligence? Saying that their peers usually know what they are talking about is no excuse.

I find your defence of descriptive sloppiness philosophically weak in the extreme.

"Just because some philosophers and lay people get confused, doesn't mean they [the scientists] need to be taken away by the language police."

What kind of argument is that? Are scientists not trying to be factually correct then? Is it not important for scientific explanations to be parsimonious? Does it not matter that philosophers understand scientists? When scientists speak of "representations" are philosophers just a bit confused about what scientists actually mean? Should we assume that the scientific usage of "representation" just means "correlation" or "handy explanatory correspondence"? Where is such an admission ever acknowledged in the literature?

"The difference is that the brain and cells embed information-processing systems within them, and thus we can speak of signals as transfers of information."

That's a new one on me! So neurons embed information-processing systems within themselves? You have to be kidding. Even a sloppy scientist would spot that false claim.

"When cells receive a certain protein on their membrane and change their behaviour according to this information - we can safely say that other cells have 'signaled' something."

No, we can't "safely say" anything of the kind. A protein is not a signal. A protein is a chain of amino acids that has a causal influence on certain cells, but a causal trigger is not a signal. Trees do not signal their leaves to drop in the autumn. Sunrise does not signal animals to wake up.

If you want to disregard the difference between causal triggers and signals then I can't stop you, but you will never convince me that no such distinction needs to be made. Nor will you convince me that there is any scientific or philosophical benefit to be gained from flouting such a distinction.

Matt said...

Language isn't everything. It doesn't even make up most of human communication. Scientific papers are full of diagrams, statistics, equations, photos, graphs, etc. Jargon terms generally refer to some understood relational heuristic in the theory; and if you want clarity, probably the language to use is information theory (often we just get away with diagrams and such), not Oxford English.

Jim Hamlyn said...

Sorry Matt, perhaps you are busy and don't have time to respond to the issues I am raising which is fine. But once again you are not presenting me with any reason to agree that careless locutions are preferable to precise ones. Let's be absolutely clear here — the issue is not with the capacity of ordinary language to describe things clearly. The issue is with the assumption that convenient active usages are justified in preference to more demanding passive usages. Why are you defending the use of the active voice in scientific discourse when the passive voice is also available? Do you really wish to add your tacit support to sloppy theorisation just because certain locutions are more difficult to formulate than others because they are less common in ordinary discourse?

Andre said...

Jim,

Signal - Miriam Webster: a detectable physical quantity or impulse (as a voltage, current, or magnetic field strength) by which messages or information can be transmitted

Jim - a signal is a form of communication in which one thing (a symbol) is used to represent another.

It seems that you are, unnecessarily, introducing symbols into your understanding of signals. Would you care to explain why you do that?

Matt said...

You're right, I am busy. Even so, ideally in science, language isn't used to theorise. Maths is.

Jim Hamlyn said...

Andre, how do you suppose that "messages and information can be transmitted" without symbols of some kind? What do think a sign is after all? Surely you are not of the incoherent opinion that signals are possible without signs?

So it is not me that is using language in an unusual way. It is anyone who takes the indefensible view that causal influences are messages.

If I transmit influenza to my son, does he get a message? Do viruses and bacteria communicate illness or do they CAUSE illness?

Jim Hamlyn said...

Matt, how does one theorise about maths, ideally?

Matt said...

Like a structuralist. ;)

Andre said...

Jim, the definition clearly states: a detectable physical quantity or impulse. So that is how you transmit 'information' without symbols.

Here is a link to where signaling is discussed: http://www.nature.com/scitable/topicpage/cell-signaling-14047077

In the article it is stated: "Cells have proteins called receptors that bind to signaling molecules and initiate a physiological response."

Could you point out your problem with a statement like this, please .....

Matt said...

http://web.stanford.edu/~montanar/RESEARCH/BOOK/partA.pdf

For something much less mathematical (though the maths above is not that difficult for the most part), check out "Physics in Mind" by a guy named Loewenstein for his discussions of information theory in physics and biology.

David said...

While I don’t share Jim’s views on language I think it is vital to look closely at the language being used by scientists to ensure it is consistently used. Scientists describe their use of terms like ‘map’, ‘information’, ‘symbol’ and ‘rule’ in a variety of different ways. Sometimes they say that they mean a term to be understood in its literal sense, sometimes they claim that the terms are homonyms with a different technical meaning than ordinary language, sometimes they say the terms are analogical extensions of the term as it is typically used, sometimes they say the terms are only used metaphorically etc. However sometimes with in a particular text they will use a single word inadvertently literally, and later in the text will switch to what can only be understand as a metaphorical use. This can lead to incoherence in the interpretation of the theory that sometimes the scientist is blind to. There is a vast literature on the use of ‘Rule’ in generative linguistics and I think that this literature has actually helped in making the science more coherent. On the issue of incoherency in Neuroscience below I have some remarks from ‘Bennett and Hacker’ which are instructive on some confusions that can occur at times (note * is placed after a word this means that the word is to be treated as a homonym with a technical meaning)
“It is extraordinarily ill-advised to multiply homonyms, but it need involve no conceptual incoherence, as long as the scientists who use these terms thus do not forget that the terms do not have their customary meaning. Unfortunately, they typically do forget this and proceed to cross the new use with the old, generating incoherence.

Ullman, defending Marr insists (perfectly correctly) that certain brain events can be viewed as representations* of depth or orientation or reflectance, that is one can correlate certain neural findings with features in the visual field (denominating the former ‘representations*’ of the latter). But it is evident that this is not at all that Marr meant. He claimed that numeral systems (roman or Arabic numerals, binary notation) are representations. However such notations have nothing to do with causal correlations, but with representational conventions. He claimed that ‘representation for shape would be a formal scheme for describing some aspects of shape, together with rules that specify how the scheme is applied to a particular shape, that a formal scheme is ‘a set of symbols with rules for putting them together’, and that ‘representation, therefore is not a foreign idea at all- we all use representations all the time. However, the notion that we can capture some aspect of reality by making a description of it using a symbol and that to do so can be useful seems to me to be a powerful and fascinating idea. But the sense in which we use ‘representations all the time’, in which representations are rule-governed symbols, and in which they are used for describing things, is the semantic sense of ‘representation’-not a new homoymical causal sense. Marr has fallen into a trap of his own making. He in effect conflates Ullman’s representations* which are causal correlates, with representations, which are symbols or symbol systems with a syntax and meaning determined by conventions.” (‘The Philosophical Foundations of Neuroscience’ p. 76)

“Similarly, it would be misleading, but otherwise innocuous, to speak of maps in the brain when what is meant is that certain features of the visual field can be mapped on to the firings of groups of cells in the ‘visual’ striate cortex. But then one cannot go on to say, as Young does, that the brain makes use of its maps in formulating its hypotheses about what is visual” ( ibid p.77)

David said...

“So, too, it would be innocuous to speak of there being symbolic representations in the brain, as long as the ‘symbolic’ has nothing to do with semantic meaning, but signifies only natural meaning (as in ‘smoke means fire’). But then one cannot go on to say, as Frisby does, that ‘there must be a symbolic description in the brain of the outside world, a description cast in symbols which stand for various aspects of the world of which sight makes us aware. For this use of ‘symbol’ is evidently semantic. For while smoke means fire, inasmuch as it is a sign of fire (an inductively correlated indication), it is not a sign for fire. Smoke rising from a distant hillside is not a description of fire cast in symbols, and the firing of neurons in the visual striate cortex is not a symbolic description of objects in the visual field, even though a neuroscientist may be able to infer facts about what is visable to an animal from his knowledge of what cells are firing in its visual striate cortex. The firing of cells in V1 may be signs of a figure with certain line orientations in the animal’s visual field, but they do not stand for anything, they are not symbols, and they do not describe anything” (ibid p. 77)

David said...

I am not saying that Hacker and Bennett are necessarily correct but I think they are correct to demand that we analyse the language used and the inferences drawn from the language very closely.

David said...

There is a debate within physics whether "shut up and calculate" is the best approach. It may be appropriate for experimental scientists but not I think for theorists. I think it is one thing to say language is not as precise as mathematics. But another thing to assert that because language is less precise than math that it doesn't matter if scientific language is confused. We wouldn't let Lacan away with confusion and vagueness so the same approach should apply to neuro scientists. Also there is no evidence I am aware of that says natural language cannot be made more precise

Jim Hamlyn said...

Thanks David, that's extremely helpful and very much in keeping with the thrust of my thinking. Inevitably it gets me branded as a "word policeman" but, as I keep repeating, I don't see why we should ignore careless locutions especially if doing so obliterates a fundamentally important distinction between causal triggers and signals, between natural processes and intentional behaviours.

I'm keen to know what Thomas makes of all this. I realise that he just thinks that bold assertions should be accompanied by hard facts. I'd like nothing more that to be able to provide some, but without wanting to seem grandiose, what evidence did Ignaz Semmelweis have at his disposal to convince doctors that their practices were too "expedient"? And how long did it take before anyone took any notice that their sloppiness was causing avoidable deaths?

Andre said...

Jim, I don't know if you missed it, but I am re-posting my previous post.

The Miriam Webster definition clearly states: a detectable physical quantity or impulse. So that is how you transmit 'information' without symbols.

Here is a link to where signaling is discussed: http://www.nature.com/.../topicpage/cell-signaling-14047077

In the article it is stated: "Cells have proteins called receptors that bind to signaling molecules and initiate a physiological response."

Could you point out your problem with a statement like this, please .....

Jim Hamlyn said...

Apologies Andre, I assumed — obviously incorrectly — that you would read David's citations of Hacker and Bennett and find your answer there.

How do you suppose that messages or information can be transmitted without symbols of some kind? What do you imagine a sign is after all? Surely you don't think that signals and signs are unrelated? I think you will find that you cannot have signals without without symbolic representation. We could discuss C.S. Peirce's misleading theory of the sign if you like but personally I prefer Hacker and Bennett much clearer version:

"But the sense in which we use ‘representations all the time’, in which representations are rule-governed symbols, and in which they are used for describing things, is the semantic sense of ‘representation’-not a new homonymical causal sense."

So it is not me that is using language in an unusual way, it is anyone who takes the indefensible view that causal influences are messages

If I "transmit" influenza to my son, does he get a message? Is the transmission of disease a "communication" in the ordinary sense of the word? Do viruses and bacteria communicate illness or do they CAUSE illness?

"Cells have proteins called receptors that bind to signaling molecules and initiate a physiological response."

The problem with the above sentence should be clear. There is no such thing as a signalling molecule unless you take signalling in the strictly causal homonymical sense and never under any circumstance unwittingly assume, nor lead anyone else to assume, that a causal trigger is a symbolic representation: an actual signal.

Thomas said...

//I'm not suggesting that scientists need to invent new terms. But nor am I convinced that "communicative expediency" constitutes justified grounds for accepting usages that come laced with further unjustified implications//
What biologist asserted these "unjustified implications"? What biologist was led astray by them? As Jan said, these are context-sensitive terms. You're imputing all these beliefs to biologists simply because they use the term 'signal'. Whatever term they used there would be a mismatch between the common usage and the exact biological meaning. What they mean by 'signal' is defined by scientific practice, not by ordinary language connotations. As I noted, your statements about signals are misguided if we take 'signal' to mean signal* in either the electrical enginering or biology sense (using Hacker and Bennett's notation). It just isn't the case that certain 'signal' involving locutions are "OK in anybody's language, whether a scientist or a layperson" unless it is clear from context what sense of 'signal' is meant. You conflate the biologists signal* with signal and have failed to address Jan's criticism that you haven't actually given any evidence of what the notion of signal* that biologists use actually is.
Do you expect the waves* of physicists to be surfable? The planes* of mathematics to be flyable? The races* of biology to have winners?

//Why would anyone who is careful in every other respect of their enquiries wish to treat the description of their findings with such negligence?//
I think they do, they treat the description of their findings with great care. But the particular labels they attach to technical notions are not very important to their findings, except as far as communicating those findings to fellow scientists goes. The findings of science are something like what experiments are done and what the results are with technical terms referring "to some understood relational heuristic in the theory" (to borrow Matt's nice account). As such, good scientists take great care in documenting methods and results, ensuring the relations and the evidence for them are clear (well, they try and of course inevitably fail, being human). They are also careful to use technical terms in ways that are clear *within the field*. They make sure the relational heuristic intended is clear. I don't really see that they should care much what particular technical terms they use or how they relate to similar ordinary language terms beyond making initial learning of the technical terms easy and disambiguating them from other contextually relevant technical terms. They care that signal* is properly understood and not confused with other technical terms, not how it relates to signal. Unless you can show some deleterious effects on biology of this re-use of 'signal', I fail to see any good argument on your part.

Thomas said...


I haven't read it, but from David's excerpts and reading some reviews/correspondence, it seems that H&B are careful to delineate the technical terms and their meanings. They don't just (using David's particular excerpt) assume that the representation of Marr and the representation* of Ullman are the same, they analyse the meaning of these two distinct usages of the term 'representation'. Also note that theirs is an internal criticism, it isn't that confusion between representation and representation* leads to confusion outside of the scientific debate but within it. They claim that 'representation' isn't used consistently within neuroscientific debate, not that the neuroscientific definition clashes with the common definition (this is complicated by the fact that one of the definitions at play is, they claim, the common definition of 'representation', but still the force of the critique is internal I take it: it isn't that this leads to confusion outside of neuroscience but within it).
As far as I can see their argument is something like:
1) The common usage of 'representation' is representation
2) This neuroscientific usage of 'representation' is representation*
3) But here this scientist conflates representation and representation* and leaps from evidence for representation* to the corollaries of representation (or something similar I gather they provide various examples of these sorts of conflations in the literature)
4) Therefore there is confusion in the neuroscientific use of 'representation'
From what I can tell this is a fine argument if it holds.
(Though I think (1) probably should be replaced with talk of another technical notion of representation, the failing is to confuse representation1 with representation2, I'm not convinced there is a commonly accepted usage of terms that carries any weight in such debates, but that opens up another big can of worms)

But, as Jan argued I think all you did was: vaguely set out (1); said nothing as to (2); nor gave any instances of (3)-style mistakes in the literature.
As such as far as I can see your argument simply amounts to disliking their usage of the 'signal' because it apparently clashes with yours: that //Maybe you want to say that. Maybe some scientist wants to say that. But I don't want to say that.// (where your disagreement seems entirely to do with their locution not the actual content, I take it you don't have any substantive issue with the existence or nature of all the various biochemical goings on captured by 'cell signalling'). So why should I preference your notion of 'signal' above theirs? You offered no argument that their accounts weren't internally consistent nor that they led to failings in biology. The best you've done is claim they don't accord with your preferred usage. They can argue exactly the same about your usage. So why should I care what term you prefer? As far as I can see (and you've given no counter evidence here) cell signalling in biology has a good track record of advancing knowledge, and the chosen term has led to no confusion beyond yours, even if it is misusing 'signal' in your view.

Aaron said...

>>: "Cells have proteins called receptors that bind to signaling molecules and initiate a physiological response."<<

In which case the signaling molecules are symbols, so you're still using symbols to transmit information.

Jim Hamlyn said...

So, what exactly are they symbols of, Aaron?

Andre said...

Jim, could you show me the correct use of signal and the incorrect use according to your understanding? I suspect you are making the definition too narrow, but I would like to see what it looks like in practice.

Aaron said...

>>So, what exactly are they symbols of, Aaron?<<

Whatever information is being transmitted.

Jim Hamlyn said...

Andre,
Signals are symbolic representations in which thing A is used to represent thing B by means of a shared rule. Language is an almost entirely symbolic system of representation. So a correct use of a signal is an agreed usage of one thing as an efficacious substitute of another.

The wrong usage of signal is any usage that assumes that a causal trigger is a signal.

Darkness causes — or at least seems to cause — countless creatures to become more passive and in many cases to slip into the depths of slumber. In ordinary language we sometimes say things like "Darkness is nature's way of telling us that it's time to sleep" or "Darkness is a signal to the body to wind down." Nobody seriously supposes that darkness literally signals or communicates with us. The relation is merely a causal one in which an absence of light triggers behavioural responses of certain kinds, the most common of which is obviously soporific.

If darkness were a signal, then an intelligent communicator must necessarily have produced it and we, the receivers of the signal, would have to know the "content" of the message. But, of course, darkness is not a signifier and nor is it a representation of any other sort. Strictly speaking darkness isn't even a causal trigger because it doesn't have any causal influence but that is another story.

Jim Hamlyn said...

Aaron, what is the difference in your view between a causal trigger and a symbolic representation?

Aaron said...

Content

Jim Hamlyn said...

So a signalling molecule has content then?

Aaron said...

Personally, I don't think information passes between cells at all, so the signaling molecules need not be symbols.

But IF information passes between cells, then clearly the symbols used to transmit these messages are the signaling molecules.

Jim Hamlyn said...

Thanks Aaron, then we needn't debate this particular point any further. If information passes between cells — which I agree it doesn't — then the cells would have to be sufficiently intelligent to understand the information.

Jim Hamlyn said...

Thomas,
"What biologist asserted these "unjustified implications"? What biologist was led astray by them?"
The important question is not what biologist was led astray but do we see any evidence that scientists and philosophers more generally are being led astray. Hacker and Bennett obviously think so and they provide compelling evidence. They would agree with me that every neuroscientist who assumes that brains generate actual representations has been led astray. Can you provide any evidence that no neuroscientist thinks there are actual representations in the brain?

You evidently think that scientists never, or at least only rarely, confuse one context of usage with another. What justification do you have to make such an assumption? Do you have any evidence that scientists do not sometimes lead other scientists and philosophers astray? Can you be sure that scientists are dully cognisant of the fact that it actually matters that they do not confuse usages?

"Whatever term they used there would be a mismatch between the common usage and the exact biological meaning."
That sounds like you believe that it doesn't really matter that scientists strive to weed out fallacious implications from their terminology because there will always be a mismatch anyway? Are you really sure that signals, messages and communication are preferable to "causal triggers", "stimuli" and "impulses"? What justification is there for creating confusion of uncertain proportions when better alternatives are available?

"You conflate the biologists signal* with signal and have failed to address Jan's criticism that you haven't actually given any evidence of what the notion of signal* that biologists use actually is."
On the contrary. Hacker and Bennett emphasise that is that it is scientists that tend to conflate usages. H&B also provide a very good explanation of the different usages of "signal" so there is no need for me to repeat them.

You say that scientists "treat the description of their findings with great care." but later you acknowledge that they are fallible like the rest of us. If they are fallible then why do you not acknowledge that they might also be mistaken in referring to causal triggers as symbols? Do you simply suppose that they are right because they are scientists? From what I can see you simply take it on faith that what scientists say is fine so long as they think they know what they mean whether or not they do actually know what they mean.

"But the particular labels they attach to technical notions are not very important to their findings, except as far as communicating those findings to fellow scientists goes."
I think you are making the mistake of assuming that there is no problem in science treating its own discourse as somehow hermetically sealed from the rest of culture.

Jim Hamlyn said...



"...good scientists take great care in documenting methods and results, ensuring the relations and the evidence for them are clear (well, they try and of course inevitably fail, being human)."
So they fail because they are human and never because they are party to any false or misleading assumptions endemic to their adopted culture of enquiry?

"They are also careful to use technical terms in ways that are clear *within the field*. They make sure the relational heuristic intended is clear. "
So when scientists communicate with people *outside the field*, they always take great care to ensure that their insider terminology isn't misinterpreted as being ordinary language then? It seems to me that Aaron has already made it clear that a straightforward interpretation of signalling molecules is that they have content. Do you want me to provide you with evidence that numerous theorists think that brains have content of this kind? Do these theorists just mean content* of a homonymical causal kind? What kind of content do you suppose that neuroscientists are talking about?

"I don't really see that they should care much what particular technical terms they use or how they relate to similar ordinary language terms beyond making initial learning of the technical terms easy and disambiguating them from other contextually relevant technical terms."
So you think that such practices are risk free and have no increased potential for conceptual confusion?

"They care that signal* is properly understood and not confused with other technical terms, not how it relates to signal."
So you are saying that they have little concern for how signal* might be confused with signal? Isn't that the whole problem? Can you provide an instance where a scientist has gone out of their way to clarify that they didn't mean signal but only signal*?

"Unless you can show some deleterious effects on biology of this re-use of 'signal', I fail to see any good argument on your part."
Unless you can show how the use of avoidable ambiguity within scientific discourse is helping the advancement of science, I fail to see any good defence of the radical carelessness that you seem so tolerant of.

Jim Hamlyn said...

So Thomas, you quite clearly see no risk of blurring or altogether obliterating the distinction between intentional behaviours (actions) and causal effects. And because you see no risk of obliterating such a distinction, you think it is ok for scientists to act as if there were no need for such a distinction either. Do you actually have a theory of the difference between actions on the one hand and cause and effect on the other? And if so, where do you draw the line between what a microorganism, cell or falling leaf does and what an intelligent creature does?
Aaron has already offered what I also take to be the case regarding the difference between causal triggers and signals: content. I can provide you with a list as long as your street of the cognitive scientists and theorists who hold that brains have content. What brains actually have is content* but it seems to have slipped your notice too that content* (i.e. causal correlations) is not content. Or do you want to argue that content* really is the same as content as it pertains to brains?

And while I am trying to expose the weaknesses in your championship of obstructive status quoism, perhaps you can indulge me by explaining what you take content to be? Do words actually _contain_ meaning? Is content something that can be measured? Is the meaning of a traffic light a property of the light? Or, is what we call "content" actually a skilful way of using one thing as a representation of another?

I just don't see any justification whatsoever for your support of scientific ignorance.

Matt said...

"Thanks Aaron, then we needn't debate this particular point any further. If information passes between cells — which I agree it doesn't — then the cells would have to be sufficiently intelligent to understand the information."

To clarify, @Jim, you're saying that cells don't communicate?

Jim Hamlyn said...

I am. I'm not sure Aaron would go so far as saying that but I certainly would, yes. You might want to take me as saying that cells communicate* but they do not communicate.

Do you want to go so far as saying that cells communicate? You can find plenty of talk of them communicating* but this is not admissible evidence of communication. If you can provide evidence of communication that is not communication* then I would be delighted because it would give me a chance to show Thomas how communication* is sometimes confused with communication.

Matt said...

Cells don't talk, therefore they don't communicate qua send/receive information? I think scientists are going to struggle to understand you.

Jim Hamlyn said...

Cells don't trade in symbolic representations Matt, or representations of any kind. Show me some information that is not representational and I will to show you that what you have is plain properties and no information.

Signals, communication and messages are intentional behaviours (actions) performed by skilled agents: perceivers.

Signals*, communication*, and messages* are sloppy scientific uses that frequently lead others (and scientists too sometimes) to suppose that cells actually produce signals, communication and messages. Why are these scientists so sloppy? Because they are blithely unaware of the need to distinguish sharply between causal processes and intentional actions.

Matt said...

Bits, not symbols. Bits.

Jim Hamlyn said...

So you are saying that a bit doesn't stand for something?

Matt said...

Of course. It's the unit of information theory. Have you studied information theory at all?

Jim Hamlyn said...

From the Wikipedia page on Information Theory: "A key measure of information is entropy, which is usually expressed by the average number of bits needed to store or communicate one symbol in a message."

Matt said...

...yeah you've got a bit of reading to do.

Matt said...

(Well, untold numbers of bits.)

Jim Hamlyn said...

What advantage do you suppose that such a juvenile rhetorical slur is going to confer upon you Matt? Can we stick to evidence and rational argument instead Please?

You are right that a bit is not intrinsically a representation but nor is it intrinsically information. The problem that you face — so far as I can see — is to distinguish between raw properties and information. If a particular protein has a causal influence upon a cell, there is no more reason to consider this an instance of information exchange than the causal influence of gravity upon a falling leaf.

From where I stand you are making the fundamental error of mistaking the pointing finger for the thing that it points at. In other words you confuse our tools for the things these tools are intended to describe.

Jim Hamlyn said...

Even if you argue, as you have elsewhere via Ontic Structural Realism and Mathematical Structuralism, that there are no relata (no objects or properties) but only “structural relations all the way down”, you still cannot use this to sidestep the distinction between the causal mechanisms that we observe amongst cells and microorganisms, and the purposefully driven actions of intelligent creatures that emerge higher up the evolutionary tree.

If the “bits” you are thinking of are in no way symbolic and are not combined to generate symbolic information then for all intents and purposes the information involved is just information* of the standard homonymic causal sort.

Andre said...

Jim, I suspect if you analyze the mechanism according to which symbols represent information, you will find it to be a similar mechanism to the one according to which electric bursts represent information to the cell.

I suspect in your understanding you are only looking at one end of the process and ignoring the role of the receptor in supplying meaning to the signal.

Aaron said...

>> I suspect if you analyze the mechanism according to which symbols represent information, you will find it to be a similar mechanism to the one according to which electric bursts represent information to the cell. <<

Nope. Such analyses have been done. No such findings have ever been produced. Quite the opposite, actually.

Aaron said...

>>I suspect in your understanding you are only looking at one end of the process and ignoring the role of the receptor in supplying meaning to the signal.<<

The cells that receive the packets of molecules that other cells send out are also unthinking cells, and thus supply no meaning.

Andre said...

Mobley, could you maybe supply a brief outline of how symbolic information differs from the way electric pulses 'carry' information?

Andre said...

Aaron, to say that cells 'supply meaning' is obviously an interpretation from the POV of the observer. The point was that a signal cannot be said to be a signal if it is not received as such.

The only reason we see it as a signal, is because it is received as one. Because of the effect it has.

Aaron said...

>> Mobley, could you maybe supply a brief outline of how symbolic information differs from the way electric pulses 'carry' information?<<

Information, by definition, informs. Packets of electrochemical energy being transferred between individual cells don't actually inform anyone of anything. That's how they differ.

>> The point was that a signal cannot be said to be a signal if it is not received as such. <<

Sure it can. If I point my flashlight into the night sky and then flick it on and off in Morse code, I'm sending a signal, but it's possible (likely, even) that no one will ever receive it. It's possible I'm just shouting into the void. But if someone DOES see my flashes (and recognizes them as Morse code and decodes them), then the signal will have been received, and a message transmitted. The SIGNAL, however, exists whether anyone receives it or not.

>>The only reason we see it as a signal, is because it is received as one. <<

Or because it was transmitted as one. Either will suffice.

Andre said...

Aaron, 'information' and 'informs' are observer judgements. It is possible to look at what happens with cells and say that the signal carries information and it informs the cell because the metaphor fits.

What is the difference between flashing randomly into the sky and using Morse code? Is it the intent of the flasher?

Aaron said...

>> It is possible to look at what happens with cells and say that the signal carries information and it informs the cell because the metaphor fits.
<<

But as you say, it's just a METAPHOR. There's no ACTUAL information because there's no ACTUAL informing going on.

>>What is the difference between flashing randomly into the sky and using Morse code? Is it the intent of the flasher?<<

Yes.
Same as the difference between typing out a message to you (which I can't be certain you will ever receive), and just pressing random keys on my keyboard. Even if something goes wrong and no one ever reads this, it's still a message, and I will still have sent it. It's not just a series of random key-presses.

Matt said...

Jim, I'm sorry you found my puns offensive. I thought they were funny; but it is clear that you *do* need to sit down and take a course on, or at least read a textbook about information theory.

I find it astonishing that there are people (plural) in this thread who claim to be in the business of clearing up confusions while demonstrating themselves to be amongst the most confused.

So the level of scientific and mathematical illiteracy in this thread indicates to me that it isn't possible to have a productive discussion here. I'm unfollowing.

Aaron said...

>>So the level of scientific and mathematical illiteracy in this thread indicates to me that it isn't possible to have a productive discussion here. I'm unfollowing.<<

I've noticed that's a general trend with both Jim and Andre'. They both seem to think that their own intuitions about how the world work are sufficient to confirm sweeping claims about the nature of reality, even though science and/or math clearly invalidates those claims.

Andre said...

Wow, Aaron and Matt, that is really condescending.

And now you will be on your hind legs defending this attitude.

This attitude is very immature. You can say what you will about Jim, and I am sure you will, but something he isn't, is immature.

Andre said...

Aaron Mobley, what is the difference between a message and just pressing random keys on your keyboard?

Can your dog tell the difference?

Aaron said...

>>Wow, Aaron and Matt, that is really condescending. <<

And also accurate. Some people deserve to be condescended to.

>>This attitude is very immature.<<

Nope. Just accurate.

>>You can say what you will about Jim, and I am sure you will, but something he isn't, is immature.<<

I've seen Jim be pretty fucking immature several times. He's pretty good about backing off as soon as someone calls him on it, but he does it quite a bit.

Andre said...

Aaron, thank you. As expected .....

Do you mind answering the question related to the thread now?

Aaron said...

>>Mobley, what is the difference between a message and just pressing random keys on your keyboard? <<

Intent, mostly.

>>Can your dog tell the difference?<<

Yes. Dogs can definitely receive and interpret messages, and distinguish them from non-communicative acts. Can she do so with perfect accuracy? No. But CAN she do so? Yes, sometimes.

Matt said...

I'm ignorant of loads of things. I just try to avoid those subjects when voicing opinions on them.

All right, in the next couple of days I'll make a post about information theory and relata in the sciences. Maybe on subsymbolic processing too if my brainmeats are performing up to scratch.

Andre said...

When one talks about signaling, you cannot ignore the context. I believe Jim is doing that.

From his perspective, the meaning of the symbol seems to reside IN the symbol. I don't think it does. The meaning resides in the context in which the symbol is created. Without the context, the symbol is meaningless.

Within the context of the cell and the signal, you have the same dynamic. If the cell does not have the right receptors for the signal molecule, then we as the observer have no way of framing the metaphor.

Jim Hamlyn said...

Aaron writes:“I've seen Jim be pretty fucking immature several times. He's pretty good about backing off as soon as someone calls him on it, but he does it quite a bit.”
When I back off it is NOT because I realise that I have been immature but because I see no point in persisting in the face of better arguments and/or evidence. I take that as one of the principal marks of maturity in anyone. Immaturity, on the other hand, is a tendency to respond to others in a childish or conceited manner and the only person I know of who regularly engages in such a self infantalising way Aaron is you. It’s a great pity, because you often have otherwise interesting or useful things to say like the following:“The cells that receive the packets of molecules that other cells send out are also unthinking cells, and thus supply no meaning.”
I agree. So am I right in assuming that you think these packets of molecules are only information* in a homonymic causal way (as outlined in the quotes by Hacker and Bennett)? Matt wasn’t clear on this distinction, so I still don’t know whether he thinks that all information is semantic or not. I suspect not. But then I don’t see why he thinks there is no possibility of confusing information* with information. It seems to me that he simply evaded the question when I quoted the Wikipedia passage on Information Theory: "A key measure of information is entropy, which is usually expressed by the average number of bits needed to store or communicate one symbol in a message." He just laid down the “I know better and you are ignorant” card and jumped ship.

Jim Hamlyn said...

Andre' writes:
“From [Jim’s] perspective, the meaning of the symbol seems to reside IN the symbol. I don't think it does. The meaning resides in the context in which the symbol is created. Without the context, the symbol is meaningless.”
I think we all agree that the “content” (the meaning) of symbols is not an attribute of symbols. The meaning of a word is not a property of the word. So NO, I don’t think the meaning is IN the symbol. Meaning is not a special attribute of things, it is a special attribute of language users. It is a socially negotiated skill in the use of thing A as an efficacious substitute for thing B. Cells and microorganisms cannot do this, which is why I deny that cells and microorganisms communicate. Cells and microorganisms communicate* (in a homonymic causal way), but they do not communicate in a semantic way (i.e. there is no meaning at the cellular or microbiological level, only at the language level).

Fred Adams said...

Quite right and there is talk of language, memory, knowledge, and cognition at these rudimentary levels if one bother to look. I agree with you that things are amiss.

Andre said...

Jim, I agree that cells and micro-organisms do not communicate in the same way we do, but I think there is enough of a similarity to warrant the description.

That is why I want you to draw me up a comparison between communication among cells and communication among people. I want to know exactly why you say there is a difference.

You are saying that what cells do cannot be called communication compared to what people do, and I want to know why you say that.

Jim Hamlyn said...

Andre', communication is a skilled activity performed by intentionally directed creatures: perceivers. Cells and microorganisms do not have skills and they are not intentionally directed because they are not perceivers. To be intentionally directed (i.e. a perceiver) one has to have at the bare minimum a capacity to produce a publicly perceptible representation of one's goal.

Andre said...

Jim, I suspect what you wrote above is the reason why you object to information transfer among cells being labeled communication.

What I am asking for is the difference in the mechanism. If you are going to claim that cells do NOT communicate, then you need to have an idea of what you mean with 'communication', so you can point to what cells are NOT doing but what we ARE doing when we communicate.

You claim that we use symbols to communicate and cells don't, but then you agree that the meaning of a symbol is not inherent in the symbol itself, but in the context of usage. The same can be said of the information cells receive. The information is inherent in the context, and not the vehicle of transmission.

So why does it matter that cells don't use symbols if symbols are merely the vehicle of communication?

Jim Hamlyn said...

Efficacious substitution. That is the difference. Communicators substitute things of one sort (representations) for things of another sort. Simpliciter

Andre said...

Communicators using symbols as vehicles for communication do. Is that the only kind of communication that counts?

Bruce said...

From an ethological perspective, communication involves the transmission of information in the form of a signal, from one individual to another. This information may be true or false but producing the signal should be of potential benefit to the individual that does it. Frankly, I don't care if it is a tree sending signals through fungal root networks, bacteria to other bacteria, cells within the body or any other life form.

The message is what an individual encodes in a signal it sends; the meaning is what another makes of it. The signal is the physical form in which the transmission from one to the other takes place. (see P.J.B. Slater’s “An Introduction to Ethology” 1985)

Matt said...

Bruce, don't bother. Scientists don't get a say here. Our job is to sit and be told how to think by philosophers fixated on idiosyncratic "ordinary language" who don't understand the tools we use. And those philosophers who do are just wrong, because they start from our mistakes, apparently.

Jim Hamlyn said...

Hi Bruce, Thanks for your comment. In reference to our past discussion, do you see why I might be uncomfortable with the notion that my "communication" of the flu virus and my communication of an idea are equivalent? Matt evidently thinks that when I communicate the flu to someone I do it deliberately. Either that or he thinks that we always know the difference between the two forms of communication involved and there is no possibility for confusion. My view is that the two processes being described are so conceptually distinct that we need to be very careful about how we use them? Is the flu information?

I have a separate challenge for Matt but I will post that as an independent thread.

Matt said...

"Deliberately" - that would be not even wrong in terms of information theory.

Matt said...

If you confuse the manifest image and the scientific image (vide Sellars) then it's simply a matter of not paying attention to (or understanding) the context.

Aaron said...

>>My view is that the two processes being described are so conceptually distinct that we need to be very careful about how we use them? <<

In other words, Matt is equivocating. That is CLEARLY not the kind of "communication" we're talking about, and to pretend like it is is to commit an error in reasoning. We're talking about the communication of PROPOSITIONAL CONTENT, not the communication of disease. Yes, we used the symbol "communication" to describe both, but they are clearly two entirely different subjects, and do conflate the two is equivocation.

Matt said...

Propositional content -- I'm a "weak" eliminative materialist, in case anyone has failed to notice. And I follow Sellars in distinguishing between the manifest and scientific images.

Matt said...

Also I'm a non-reductionist. Following Ross and Ladyman's reworked (ie, eliminating the abstracta/illata distinction) and information-theoretic take on Dennett's notion of Real Patterns (I realise that was in some ways a reposte to EM but my EM accommodates that).

Aaron said...

Then the topic is the kind of communication that is ALLEGED to have propositional content, which is still distinct from communication of communicable disease. Thus, there's still an equivocation.

Matt said...

That's like saying that constants, variables and coefficients should be called different things in different disciplines that make use of them.

Matt said...

And yeah, propositional content is in the communicated data, detected by information processing structures in your nervous system. I mean, linguistics/psycholinguistics uses information theory too.

Thomas said...

I think Aaron is generally right in his //That is CLEARLY not the kind of "communication" we're talking about// comment. As I think is Matt, I don't think Matt was equivocating here, Jim was. The notion of communication the biologist is talking about in regards to disease transmission (or similarly in cell signalling) is clearly not the kind of communication spoken of in regards to propositional content (with similar arguments to be made about information and the other things that came in). As Aaron said, failing to distinguish the two is a case of clear equivocation. Yes, the term 'communication' is used to describe both but I think any decent familiarity with the literature reveals these are different uses of the term.
Jim seems to think the biologists use of the term 'signalling' or 'communication' is fundamentally misleading so they should stop this radical carelessness. I see no good argument for that. The biologist could equally argue the same of Jim's use. The biologist has a perfectly serviceable scientific theory (AFAICT) to back them up on their use of the term 'cell signalling', a term which seems to serve them perfectly well. I asked Jim for a counter here, he gave none (I'm less sure Jim has a good theory to back him up, but I'll assume he does here).
Instead of actually criticising their usage Jim seems to think the mere possibility someone else could be misled to think they meant the propositional content notion of communication means the biologists are trying to wantonly deceive. I think decent familiarity with both bodies of theory makes the distinction clear and the only radical carelessness is Jim failing to understand the two distinct uses and so equivocating the two (plus any other scientist or philosopher who became misled without actually understanding the source material).

Matt said...

Just google

"information theory"+biology

Then replace biology with linguistics. And then whatever. Same information theory, applied in different scientific theories. Kind of like, well, calculus, linear algebra, etc.

I just happen to think that the world is the totality of non-redundant statistics (modelled by mathematical structures), not of things. Information theory is the most obvious tool to use and it is, as far as I know, available to any scientist.

Thomas said...

Indeed, information theory applies to all. Though then it gets even more complicated to speak of. Given the troubles so far I don't expect much headway here.
I don't think the propositional attitudes philosophers speak of are instances of Shannon information but they can of course be analysed in terms of it (or something like that, as I said hard to speak properly here). Just like Jim's falling leaf can be analysed in terms of Shannon information (if information theoretic accounts of physics are true). Plus propositional content can be analysed in terms of Shannon information at multiple levels. We can consider the Shannon information of sender-message-receiver, or of the causal processes that realise the utterance (and don't think there is any simple linkage between these - don't think causal processing involving x bits need result in a message of x bits, or f(x) bits).

Matt said...

I assume this is done, then?

Jim Hamlyn said...

Sadly Matt doesn't wish to provide an explanation of intentional behaviour via Information Theory or to say at what point a creature might qualify as an agent as opposed to a behaviouristic responder. Certainly nobody believes that cells or organs are agents but this doesn't prevent many scientists and theorists form mistakenly attributing agency to brains. Whether Matt takes any steps to avoid such conceptual confusions in his own discussion of brains is not sufficiently clear to me but what is clear is that he sees no conflict in attributing the terminology of intentional acts — of agency — to cells, so I doubt that he sees any significant issue in in ascribing psychological predicates to brains.

As regards Information Theory. I have no issue with any tool that helps to represent the universe. But what I object to is the assumption that the tool and the thing it _describes_ are one and the same. The universe is neither a representation nor a bunch of representations but It seems to me that Matt is so beguiled by Information Theory that he mistakes the pointing finger for the thing it points to. He claims the universe is representations (data) all the way down. I think he is wrong. Whatever it turns out to be, the universe or its fundamental parts will never be representations because representations are the preserve of skilful organisms alone and are produced for the explicit _purposes_ of communication.

Matt said...

I haven't provided an Information-theoretic account of intentional action? Parallel Distributed Processing - it's information being processed. This isn't difficult.

Matt said...

Jeeze every model of cognition there is uses information theory at least tacitly. If you're struggling with information processing then that's your problem.

Matt said...

http://journals.plos.org/ploscompbiol/article?id=10.1371%2Fjournal.pcbi.1003588

Matt said...

It this point, Jim, all you're doing is exhibiting profound scientific illiteracy.

Jim Hamlyn said...

Who is the more illiterate I wonder: the person who seeks conceptual clarity across domains or the person who does not and resorts to insults when challenged to justify their reasoning?

Matt said...

I justified my reasoning with reference to science as it is. The insults (if there are any) came for free.

Thomas said...

"the person who seeks conceptual clarity across domains"
But you can't find any meaningful conceptual clarity if you don't look at the actual concepts that are used in the domains. If all you do is equivocate every use of the term 'signal' (say) across all domains then you're not promoting conceptual clarity just committing a gross mistake.
At best you are committing what Dennett in his response to H&B terms "conceptual myopia: treating _one's own_ (possibly narrow and ill-informed) concepts as binding on others with different agendas and training". At worst you are completely failing to engage with the domains and equivocating among entirely different (if probably related) concepts. I remain unsure you are only doing the better of the two.

Matt said...

Also, I'm not sure how I'm mistaking data with "the thing data is supposed to represent". Data interacting with sensory systems and cells in your body is literally everything in your experience, whether linguistic or sublinguistic, phenomenal or nonphenomenal, whatever. By definition, you can't verify or even experience the reality of anything beyond data. If this doesn't make sense, again, research information theory.

Thomas said...

Jim: "[Matt] claims the universe is representations (data) all the way down. I think he is wrong."
You're equivocating information (in the Shannon sense) and representations (and data). Given Matt's reference and even wiki seems too hard for you maybe try James Gleick: "Shannon said that the notion of information has nothing to do with meaning. A string of bits has a quantity, whether it represents something that’s true, something that’s utterly false, or something that’s just meaningless nonsense" or represents nothing at all (http://www.wired.com/2011/02/mf_gleick_qa/).

There is actually a theory in physics that in fact something very much like 'the universe is information all the way down' holds (with varying interpretations), digital physics as it is often called (http://en.wikipedia.org/wiki/Digital_physics). As John Wheeler says "all things physical are information-theoretic in origin" or as Chalmers says "a conception of the world on which information is truly fundamental". So information may be in some real sense both the finger and the thing pointed to.

Why the Basis of the Universe Isn't Matter or Energy—It's Data | WIRED
James Gleick's new book documents the rising role of...
WIRED.COM|BY KEVIN KELLY

Matt said...

Thomas, you'll appreciate this: http://arxiv.org/abs/quant-ph/9605039

That's actually my preferred interpretation.

[quant-ph/9605039] Quantum Information Theory of Entanglement...
ARXIV.ORG

Bruce said...

There is in my mind no doubt that information theory is fruitful but one should not discount that information can be considered as a non-random structure, and as emergent whether or not Shannon's model can comfortably acommodate this. We must also resist the temptation to view the universe as a product of Laplace's demon - deterministic. It is a world of propensities and changing propensities that have probabilities ranging between 0 (impossibility) to 1 (certainty), not just only 0 and 1. Propensities, like fields of forces, suck us forward. So yes information theory is fruitful, but this does not mean that a reductive, non-emergent interpretation is adequate.

Jim Hamlyn said...

"By definition, you can't verify or even experience the reality of anything beyond data."

Can I kick data? Do I have data for breakfast in the morning?

"Data interacting with sensory systems and cells in your body is literally everything in your experience..."

So I'm data interacting with data then?

Even if I accept your ontology, we are still left with the risky business of not clearly distinguishing between mere behaviours and intentional actions. In a forthcoming contribution to a book edited by Luciano Floridi, Fred Adams (who commented earlier in this thread in support of my position) writes:

"Shapiro (2007) points out that E-coli have the capacity to detect the difference between glucose and lactose and turn on or off the systems internal to them that chemically process the different substances. What is more, the mechanism by which they do this can be represented as computational steps."
[...]
"This kind of explanation rises above the level of physics alone, chemistry alone, and biology alone. It does so because it involves reference to information. So the obvious question is what is the next level of explanation going upward from biology? The next science up is typically psychology and cognition involves a type of psychological state/process that exploits information. So it is quite understandable why scientists and philosophers may believe that it is cognition that is exhibited by bacteria (or plants)." Fred Adams, "Information and Cognition" (forthcoming) in L. Floridi. Ed. volume on philosophy of Information.

Fred said...

Hey those are my words!

Jim Hamlyn said...

Damn good ones too. Is that ok though - I can remove them if you like?

Fred said...

You don't need to remove but do identify the source please
This will be published soon but I don't want to be scooped before it appears

Jim Hamlyn said...

Done. And thank you.

Bruce said...

"So it is quite understandable why scientists and philosophers may believe that it is cognition that is exhibited by bacteria (or plants)" - F. Adams (quoted from above)

Jim

Of course it is obvious that I do not say that - cognition as we consider it in higher animals is not necessary for knowledge at lower phylogenetic levels.

Jim Hamlyn said...

Bruce, I would argue that knowledge -- i.e. the capacity to represent our perceptually acquired skills -- is the very mark of cognition.

Bruce said...

Jim, your view what what knowledge does is restricted by your idea of what knowledge is. What is questions rarely lead to satisfactory answers to our problems.

Jim Hamlyn said...

You are welcome to take that view of course Bruce, and I respect your right to do so. I guess my challenge would be to ask what quantum of cognition can you possibly allude to, by whatever direct or roundabout way you wish, to unsettle my conviction that knowledge is no more and no less than skill?

Bruce said...

I agree with Karl Popper that expectations (largely unconscious in evolutionary history) are homologous to knowledge. Propensities that may or may not be realised e.g. the genome, the dog's fretting that his master has not returned home, the trees propensity to drop leaves at autumn.

Matt said...

"Can you kick data?" can you smell wormholes?

Jim, you're just flat out wrong, stuck in the manifest image. Which I suppose is comfortable because then you don't have to bother with science.

Bruce: on pure determinism, an information theoretic ontology does not give you that except where it actually occurs in the special sciences; and it clearly isn't a fundamental property of reality. See Ladyman, Ross et al. "Every Thing Must Go".

And on that note, I never said I was a reductionist. I'm not, and I don't know of any structural realists who are.

Patterns of data are scale-relative and they rarely reduce to others. Each special science deals with certain scales. There's no global supervenience here; just permissibility ultimately from fundamental physics and often other special sciences. Being a structuralist, I'd say that theories of different scales can't reduce because but they track novel patterns of data. There's maybe cases of local supervenience in the brain sciences, but these are only confirmed where the structures of the two disciplines are shown to be at least homomorphic.

Matt said...

Basically the scale relativity of disciplines and so theories that I was referring to (that and the demarcation between the scientific and manifest images), we have our "clarification" issues sorted out. Every locator in a theory is ultimately relational, regardless of what you call it, and if you understand the structure of the theory (which most practicing scientists do; and neural networks have been trained to, by Thagard and others), then you can communicate with your colleagues without confusion. Sure neoscholastic philosophers who know little of the discipline will be left scratching their heads but really, who cares.

Bruce said...

Matt, thanks for the article you posted. I find a lot of resonance with David Deutsch's "The Beginning of Infinity Explanations that Transform the World" (2011), and although our sets may not overlap entirely I think there is great overlap and fruitfulness to be uncovered.

Jim Hamlyn said...

"Jim, you're just flat out wrong..."

On the question of the nature of reality, I may be flat out wrong, but then again so may you. I haven't even offered a theory of reality, just as you have not offered a theory of the necessary and sufficient conditions for intentional directedness.

Saying that someone is flat out wrong when there are obviously two completely different issues on the table, both of which are contested, is either profoundly naive or profoundly ignorant. It certainly doesn't reflect well on your capacity to disambiguate more finely woven issues.

"Also, I'm not sure how I'm mistaking data with 'the thing data is supposed to represent'".

Firstly because you conflate data* with data and secondly because you refuse to acknowledge the importance of disambiguating the two.

Matt said...

I conflate "data with data", I think you just failed by your own criteria, but because I understand context, I know what you're saying.

You mean to differentiate between the way that the detectors in the LHC detect leptons by acting as receivers with the way that, say, photoreceptors act as receivers for photons? I argue that in both cases, the detections cause a cascade of communication through processing mechanisms. Notice that in those cases I can describe things using information theory alone. According to the data, the language is unnessary for mapping the fundamental ontology of the patterns being studied.

When both cases are clearly most parsimoniously described as information theoretic and the nouns and verbs merely works as an imprecise but clearly useful heuristics for locators and information-theoretic projections in an empirically precise way? Similar to how computation as the thing being studied co-occurs with many other modal structures with corresponding special science (and the fundamental structure of quantum physics)? Or how those structures are all themselves precisely tracked information-theoretically? Structural realism goes no further than the data and treats the locators as heuristics. Scientists rarely get confused because they know the relations that comprise the theories. If they don't then they will struggle for favourable evaluations from their peers.

As I said, sometimes we need better heuristics, probably moreso in multidisciplinary sciences, but you need to know the shit out of these sciences to do this. In fact, information theory here serves as the criteria for deciding whether locators at different scales are likely to be confused by elucidating the differences between the data patterns they track, but again this will only be useful to multidisciplinary sciences. An example of this being done successfully is bits vs qubits; there was potential for confusion in some first year computer science and quantum mechanics double major undergraduates (I don't see how given that the bits in both cases are very clearly embedded in different structures), but it was easy to do. Another difference is particles in classical mechanics and those in quantum mechanics - nobody confuses these two heuristics, because again the structures they (relatively crudely, compared to information theory) track, are starkly different (despite the findings I linked above). Sometimes you hear "quantum particles", but it's rare, because it's unnecessary.

So, to summarise: you can't differentiate such heuristics in complicated cases (such as those in the brain and cognitive sciences) with intuition alone, you need to do it with an intimate familiarity with the structures those sciences track.

Jim Hamlyn said...

_________________________________________________________________


The following are extracts from a discussion on Facebook on the same subject and involving several of the same people.

_________________________________________________________________

Gottlob said...

What is the definition of information?

Bruce said...

"Information is a non-random structure or pattern of relationships within a system, indicating future interactive potentialities. It either originates along with the system, or is acquired or developed by it in the course of its interacting with and responding to its environment, and the problems generated by that interaction. Note that this definition does not require correspondence between information and the environment. Nor does it assert that information is encoded in some simple cause-and-effect fashion, but leaves room for emergent information in the context of a system's interaction with the environment." Joseph Firestone

http://www.kmci.org/media/Whatknowledgeis%20%28non-fiction%20version%29.pdf

Matt said...

I will say this though: "What is the relationship between information and a physical substratum?"

What is the evidence that these two things are distinct in the scientific image? Ha, there is none.

Jim Hamlyn said...

Bruce quotes Joseph Firestone: "Information is a non-random structure or pattern of relationships within a system, indicating future interactive potentialities."

Indicating is a skilful activity that only symbol users can understand. Indications are not objective attributes of things, they are learned techniques for interpreting things in symbolic terms. Furthermore, only the most intelligent creatures are capable of discerning "future interactive potentialities".
This is precisely the kind of careless usage that I have been talking about. It fails to distinguish between causal homonymic usages and semantic usages.

Thomas said...

Jim: // Bruce quotes Joseph Firestone [...] This is precisely the kind of careless usage that I have been talking about. It fails to distinguish between causal homonymic usages and semantic usages.//

If you'd actually looked at the source (Riskonomics: Reducing Risk by Killing Your Worst Ideas, p.4) you would have found a few pages down:
//You may have noticed that the ideas of 'information' as a non-random structure, and as 'emergent,' don't fit well with the account of information that, by far, is the most successful, namely Claude Shannon's account. But, it is important to understand that Shannon's model is, and from the beginning was, intended to be an account of information that focused on only one aspect of the pre-existing notion of 'information' common in ordinary language, specifically the gain in information resulting from a communication of symbols or bits.//
The source is clearly talking about what you term the semantic kind of information or something similar, and clearly distinguishes his usage from the Shannon kind.
This is precisely the kind of careless criticism I have been talking about. Failing to distinguish between what authors actually mean by terms and your imputed meanings (Dennett's conceptual myopia).
[Not to endorse Firestone's account, I only skimmed a few pages, looks like he may be synthesising the two views to some extent, don't know I agree]

Thomas said...

To be clear Jim, I'm not criticising you for not having read the piece quoted. It's fine to just deal with the excerpts posted (Brian expressed a view through the quote separate from the authors view). I was just using this trivial example to reiterate my point that when accusing people of equivocating concepts it is critical to establish what concepts they are actually using. Just looking at what terms they use won't give you the concepts, you have to look at the usage of the term. Having been clear that terms (/symbols) aren't the vehicles of concepts I'm sure you appreciate this.

Gottlob said...

Indicating does not necessarily mean an intelligence is indicating; dark clouds in the sky indicate rain, or a drought indicates a possible upcoming famine.

Jim Hamlyn said...

Fair enough Thomas, point taken. If I had Firestone's book I would certainly have scanned it for context before mouthing off. Nonetheless it seems pretty clear to me that Firestone is making the mistake of taking the Information Theory usage too far and imputing intentionality (in both senses) where there is only causality.

John R said...

How about this? An intelligent agent understands that dark clouds are an indication of rain.

Jim Hamlyn said...

Indications are ascriptions. The indication "in" a rain cloud is a skilful attribute of us Gottlob, it is a way of using the perceptible characteristics of the cloud as tools for the prediction of a future state of affairs.

John R said...

Jim, most animals can perceive dark clouds and often make preparations for rain. Could you explain what they are perceiving "in" the dark clouds?

Thomas said...

Jim: "If I had Firestone's book I would certainly have scanned it for context before mouthing off"
Bill posted a link to an excerpt including the bits he and I cited directly under the selection you quoted - http://www.kmci.org/.../Whatknowledgeis%20%28non-fiction..., p.4.

"Nonetheless it seems pretty clear to me that Firestone is making the mistake of taking the Information Theory usage too far and imputing intentionality (in both senses) where there is only causality."
From Firestone: "For example, you create beliefs (information in your mind) that, to the best of current knowledge, can’t be accounted for by the causal interaction of your brain with your environment."
I think Firestone is quite clear here that the sort of information he is talking about is not causal and cannot be reduced to the causal (again not to agree with his emergentist line). He quite clearly distinguishes his notion of information from Shannon information and from any direct causal account. From my quick skim. Care to actually give any evidence from Firestone about what he thinks rather than just making, as far as I can see, unjustified assumptions.

Though again my key issue is not about you failing to read that one piece (or even you failing to follow the link directly below the bit you quoted and then claiming you didn't have access to the piece to scan). Its about you failing to read the works you want to include under your 'conflating intentional behaviour with causal processes' (or something like that) banner. Did you scan any (much less every) neuroscience article that uses the term 'representation' for context before saying "every neuroscientist who assumes that brains generate actual representations has been led astray"? Certainly a lot of neuroscientists use the term 'representation' (in some sense). Did you actually check it was anything like your sense so that your criticism of 'neural representation' (in your sense) actually applied?

Jim Hamlyn said...

Yes, we've discussed this before haven't we John (quite a while ago though)? Nonverbals — cows say — tend to move towards shelter when the weather becomes overcast. Some theorists assume that the cows are making an inference about the weather but, as you know, I think that is way way too extravagant. To make an inference we first have to be able to perform logic operations on abstract non-particulars. So how do our bovine friends whose brains we can cup in our hand manage to pull off the feat of clairvoyance necessary to respond efficaciously to overcast weather? They learn to — either from other cows or through repeated exposure to regularly correlated conditions.

Jim Hamlyn said...

Thomas, Whilst I freely admit that I overlooked the Firestone link I fail to see how this undermines my claim that every neuroscientist or philosopher (including Firestone) who assumes that brains generate ACTUAL representations has been led astray. Some neuroscientists and philosophers clearly do hold that brains _contain_ representations — not just causal correlations but actual representational stand-ins. As I said, Hacker and Bennett would agree with me that these neuroscientists and philosophers have been led astray. Representations are intentional artefacts and behaviours, they are produced by perceivers for the purposes of public communication with other perceivers. There aren't any perceivers within organisms let alone organs or cells.

Bruce said...

But Jim you are not acknowledging that even at the Precambrian cellular level it is the active chemistry and integrated organelle/ cellular activity that pushes the organism into problem situations. These are resolved by the death of the organism or by mutations successfully acommodating the challenges. There are even active epigenetic processes. You seem to want to draw a line beneath a word "perception" and then mount an argument around the word. "Knowledge" at any phylogenetic stage can only grow by trial and error or criticism of existing propensities. Existing knowledge can be amalgamated but this reinforces dogmatism or inertia, it is not growth.

John S said...

Its very simple folks. If you really believe that Knowledge can be reduced to formal mathematical, logical and Scientific models,.then there really is no point of positing anything on this OP. Because if you do disagree with me you are doing it as an Intelligent agent, which refutes your position. This does not mean that intelligent agents are not made up of physical stuff.

Jim Hamlyn said...

Bravo John S! I hadn't thought of putting it like that. We could put it like this too:

If you believe that Knowledge can be reduced to symbolic representations (language, logic, mathematics) then there really is no point of positing anything on this OP. Because if you do disagree you are doing it as perceiver possessed of nonverbal skills which refute your position.

Bruce said...

Information can be incorporated in the discussion without relapsing into reduction.

Jim Hamlyn said...

Bruce, I agree that there is information on my cereal packet but I do not agree that the cereal is information.

Bruce said...

"Bruce, I agree that there is information on my cereal packet but I do not agree that the cereal is information." Jim Hamlyn

There are at least 3 issues:

1. The signals typed on the cereal pack contain messages from the owner of the product to the consumers. They are intended to inform the consumers.

2. The cereal is information - structures in the universe that can be handled/ eaten or whatever by us.

3. What we conjecturally declare about the pack and the contents, knowledge - a subset of information.

I don't know why you said the above, Jim.

Jim Hamlyn said...

Bruce, John S already put it very well I think. The tendency to describe everything as information is a mistake of intellectualists who suppose that the skilful use of symbols and the world these symbols describe are indiscriminable. So, for example, Matt responds to Gottlob's comment:

\\'What is the relationship between information and a physical substratum?'

What is the evidence that these two things are distinct in the scientific image? Ha, there is none.//

Cereal is only information to those capable of treating it as information; i.e. informed perceivers adept in the use of symbols.

Bruce said...

I agree with Karl Popper when he says "Thus there is no uninterpreted empirical basis; and the test statements which form the empirical basis cannot be statements expressing uninterpreted 'data' (since no such data exist) but are, simply, statements which state observable simple facts about our physical environment. They are, of course, facts interpreted in the light of theories; they are soaked in theory, as it were." ("Conjectures and Refutations")

But our interpretation of the world is that there is a reality separate from our understanding of it, and we are usually prepared to accept that our notions of cereal packets and information models reflect reality well enough compared with other notions to the contrary.

Jim Hamlyn said...

I like that quote from Popper a hell of a lot Bruce. Thank you.

I suspect that Matt would say that there ARE uninterpreted data and science is in the business of interpreting them. If he wouldn't then our disagreement is over nothing.

"But our interpretation of the world is that there is a reality separate from our understanding of it..."

Our _understandingS_ of it, yes, absolutely. At last we have reached some significant agreement. Braw!

The reason I am so adamant that the intellectualist is mistaken is because one of our repertoires of understanding -- arguably the most fundamental -- is by way of nonverbal representation. You even used the metaphor of "reflecting" reality after all which is duplicative in a way that language simply isn't.

Whether we construct a model, a description or an image we are in the business of producing representations for the purposes of negotiating our interpretation of the world. Data, information and truth are each part of that process of interpretation but they are not the only tools we have. I am urging you to notice that data, information and truth are fundamentally verbal in nature whereas the reflection in a mirror is not. A reflection of "uninterpreted data" is not by virtue of reflection an interpretation. Do you see what I am getting at?

I'm suspicious of the word "interpretation" though because it seems weighted towards verbal skills. When I balance an object is it an _interpretation_ that one side is equal to the other in one fundamental respect? I don't think it is. But nor is it soaked in theory. It is soaked in knowhow.

Gottlob said...

Jim,

Information is just a logical relation. It doesnt require an interpreter.

Also, does indicating something always have reference to an intent? Like if I say flowers indicate the coming of spring, is there an intent in the way I am using the terms flowers and spring?

Information I think doesnt mean a self conscious attribute of nature. It is just a technical term for a substrate that is having a certain type of cause and effect relationship. Like the information in a turbulent stream of fluid may be the characteristics its heat and momentum have, and how they interact to form some general law of flow of fluids.

Jim Hamlyn said...

Yes Gottlob, I think I get that. Could you explain though what you see as preferable about such terminology over vanilla terms such as "objects", "properties", "substrates" and "characteristics"? How is this technical usage of "information" more illuminating? Surely "information" is the verbal label we use to grapple with the substrate in abstract terms?

Also, are you sure that these informations "interact to form a law"? Would it not be more accurate to say that they interact in regular law-like and thus efficaciously predictable and exploitable ways?

Indicating is a skill both of production and reception. To be capable of producing indications is to be capable of interpreting indications. To be able to point is to be able to understand the skill of pointing. All skills are intentional, yes.

Gottlob said...

Jim; Vanilla terms, lol. By properties, substrates and characteristics I am meaning the aspects of nature that physics studies. I donot know what you call them. Like flux, capacitance, ambience, laminar flow, etc. these are terms for certain physical characteristics of nature, and from studying them we get photons, atoms, quarks, black holes, etc. Laws just describe how these things tend to behave. From the term information what I gather is it is just giving the logical relations within these physical parameters. Logic and physics are similar but different things as the latter is empirical whereas the former is not. I think information is just describing the logical forms in physics or biology. I dont think Matt is using it in the sense of a Buddhist self aware metaphysical attribute, or anything new age or of that sort.

John S said...

Matt, I don't have any problem with Cognitive Science at all. I only have a problem with your Philosophic conclusions about the interpretation of its findings. We call guys like you followers of Scientism. That is the faith that Science alone is the arbiter of all that is Real. While Science as opposed to Faith, Authority, and Brute force may be the most useful way of settling our beliefs of what is real, you must remember that Science itself grew out of the human experience of trial and error. This is known as the Ethics of Belief. In other words Ethics is prior to Science. And Ethics requires the choices of Persons that are Agents.

Jim Hamlyn said...

"I dont think Matt is using it in the sense of a Buddhist self aware metaphysical attribute, or anything new age or of that sort."

No, certainly he is not... to his great credit.

He holds that there are no things, no objects only relations. At other times he has claimed that there is only data. What would clear things up for me, regarding his ontology at least, is whether he thinks relata are the substrate, so to speak, to which the data accurately refer (in the ways you clearly delineate) or whether he thinks relata and data are one and the same.
Anyway, that is a secondary issue as far as I am concerned. The principal issue for me is intentional directedness: action. I don't think you can get intentional directedness out of data but I'm happy enough to accept that we can get it out of relata -- so long as these are sufficient to constitute perceivers: i.e. agents capable of producing perceptible representations of the things with which they are engaged.

Thomas said...

Gottlob: //Like the information in a turbulent stream of fluid may be the characteristics its heat and momentum have, and how they interact to form some general law of flow of fluids.//
Jim: //Also, are you sure that these informations "interact to form a law"?//
I don't think Gottlob said that at all. Though given the limited precision of ordinary language, and human language users (not to single out Gottlob), it is a bit hard to tell. I'm pretty sure he was saying that information is about various physical characteristics and how these characteristics interact to form a law. Not that the information itself interacts to form a law. Or are you conflating the pointing finger and the thing pointed to now Jim? I generally agree with Gottlob here but don't agree at all with your reading.

For someone who seems to emphasise (often to exclusivity) the importance of ordinary language you don't always seem very good at understanding it Jim, IMHO. Perhaps you should work on your critical reading skills (as many of my previous rejoinders have pointed to). You should probably also accept Matt's view that ordinary language is a very inexact tool and perhaps thus familiarise yourself with better ones like formal mathematics and logic and the technical concepts they support. That would seem to help your apparent search for exactitude.
At least you could try and cede to those who actually understand such formal tools and terms, or secondary sources better placed than you to do so (as I try to do, not being that familiar with many of the formal tools and concepts myself). Rather than just wantonly interpreting their work solely on the basis of inexact and idiosyncratic ordinary language accounts intended to be understood by peers in light of actual understanding of the formal methods and concepts.

You certainly shouldn't continue to paint the views of these formal tool users solely in your basic ordinary language colours. To wit:
//How is this technical usage of "information" more illuminating? Surely "information" is the verbal label we use to grapple with the substrate in abstract terms?//
No, information in the technical sense is an arbitrary verbal label for a well-defined mathematical concept. Yes, used for something like 'grappling with the substrate in abstract terms' but not defined in such a simplistic way. This technical definition differs markedly from the ordinary language concepts of //"objects", "properties", "substrates" and "characteristics"// and any corresponding technical concepts of those. Plus (to I think re-iterate Gottlob's "Vanilla terms, lol") none of those concepts are clear in either their ordinary language or technical usages or in the relation between those.

John S said...

I think, as Sellars said, thinking is a form of covert speech. If he is right all of our thinking in this OP involves language and language needs at least two intentional agents to work.

Jim Hamlyn said...

Thomas, I asked Gottlob the question because it wasn't clear to me what he thought. He provided a very clear answer that I then acknowledged as "clearly delineated".

In another recent thread Matt wrote:

"I just happen to think that the world is the totality of non-redundant statistics (modelled by mathematical structures), not of things. Information theory is the most obvious tool to use and it is, as far as I know, available to any scientist."

Once again it is not clear to me what Matt is saying, because on my reading he is saying that the world actually IS the totality of non-redundant statistics, not things. Are you saying that I am misinterpreting him? Is his point that the world is the totality of non-redundant relata that can be statistically modelled by mathematical structures? I have no problem with that reading but I am not certain that it is what he actually believes.

Your response to him was to say "Indeed, information theory applies to all." That suggests to me that you took the "applies to" construal, not the construal I am questioning. Nonetheless you obviously think that information theory also applies to skilful action and the nonverbal capacities of representation that enable much of it. That makes you an intellectualist doesn't it? Or am I misinterpreting you too?

Jim Hamlyn said...

Furthermore Thomas, and more to the point of my original enquiries, do you take the purposeful actions of perceivers (actions of representing, in particular) to require within the body of the representer some currently triggerable disposition to represent an anticipated goal or purpose of the currently ongoing representing action?

If yes, then where in a cell is there an an embodiment of a triggerable disposition to represent this cell's anticipated consequence of the action of purposefully offering a representation to another cell? How do we get it to demonstrate that it 'knows' what it's doing?

If no, then is it only cells and other lowly organisms that are incapable of purposeful action, or is there 'really' no such thing as a purposeful action of representing? Not even we ourselves can 'really' perform the purposeful action of signalling to our friend that we are joking by winking as we speak, and explain to somebody else that this what we're doing as, or even before, we do it. Are you an old style radical philosophical determinist under that veneer of analytic sophistication? Do you suppose that we are just deluded (poor mugs!) when we think that our purposeful actions (notably, actions of representing) differ significantly from our mechanically propelled, inescapable, pratfalls on ice?

Thomas said...

//Thomas, I asked Gottlob the question because it wasn't clear to me what he thought. He provided a very clear answer that I then acknowledged as "clearly delineated".//
Did you? Where? I must be missing that. I can't see you say anything like that. The only reply I can see from you dealing with the post I quoted from is:
//"I dont think Matt is using it in the sense of a Buddhist self aware metaphysical attribute, or anything new age or of that sort."

No, certainly he is not... to his great credit.// and then some other shit about what you think Matt means by terms.

Where was this acknowledgement? What were you acknowledging? So you now realise that the technical notion of 'information' is better than //vanilla terms such as "objects", "properties", "substrates" and "characteristics"// as you seem to have previously held?
Or are you just going to keep dodging the key points of my comments?

Jim Hamlyn said...

Yes, you obviously missed it. I wrote: "What would clear things up for me, regarding his ontology at least, is whether he thinks relata are the substrate, so to speak, to which the data accurately refer (IN THE WAYS YOU CLEARLY DELINEATE) or whether he thinks relata and data are one and the same."

I have acknowledged the usefulness of your comments several times in the past Thomas as you well know. Very rarely, if ever, have you addressed those I have put directly to you.

No I don't accept that the technical usage of information is preferable to standard talk of "objects", "properties", "substrates" and "characteristics". As I keep saying, and you keep avoiding, I think it obliterates the distinction between behaviours and actions. Perhaps you should read some recent papers by Fred Adams (who commented in support of my two most recent posts) and then you might notice that there may be some method to my madness. Either that or Adams is mistaken. I'll take my chances with his _informed_ opinion.

Thomas said...

"Yes, you obviously missed it. I wrote..."
That was a reply to Matt's views not Gottlob's. The comments I posted were from Gottlob, it was his question I claimed you failed to address not Matt's, as I think I made clear. Or are you now equivocating their two views? Got some evidence they mean exactly the same thing?
//No I don't accept that the technical usage of information is preferable to standard talk of "objects", "properties", "substrates" and "characteristics"//
So you didn't actually address the question *of Gottlob* I pointed to.

"then you might notice that there may be some method to my madness. Either that or Adams is mistaken. I'll take my chances."
Wow, so either you are right or Adams is mistaken because he gave some support for your views? Sorry that doesn't work. Not to attack Adams, I won't equivocate your views with his even if you want to. Though I will look at Adams' views to see if I agree with what he actually says. Maybe you could also look at the actual views people express rather than just lumping them together (Matt+Gottlob) or imputing views to people you didn't even read (Firestone).

"I have acknowledged the usefulness of your comments several times in the past Thomas as you well know. Very rarely, if ever, have you addressed those I have put directly to you."
Yeah, and I am *kinda* sorry for that (though will continue to do the same). I have dropped some of our discussions due to other commitments, or lack of a well-formed reply, or finally having a well-formed reply when the thread had moved on. I have drafted a few replies to you I never posted. While not posted those inform my current replies. I won't criticise you for not replying, I will criticise you for what you do reply, I hope you do the same.
Though i would note you are rather selective in your responses. You often ignore the brunt of replies to pick up on side issues (i.e. here where you failed to address anything but my first sentence), IMO of course. So yeah, we are both, like most authors, selective in replying only when we have a good rejoinder.

Though given you seem, repeatedly, to be unable to generate well-formed replies to the actual positions of the authors you attack I'm rather unconvinced that giving a positive reply of my own to your challenges would achieve much. Perhaps if you stick to comments on other authors that actually attack *their view* and don't invite a simplistic counter about you failing to engage I might actually have to deal with your attack and really proffer my own views.

Bruce said...

"I think it obliterates the distinction between behaviours and actions." Jim Hamlyn

This is the mystery to me. I am the first to acknowledge that higher capacities of language have enriched our propensity for behavioural plasticity, and we are actors - in the sense of having some quality of consciousness and symbolic goal-directedness.

But even though paramecia are not conscious actors, nor manipulators of symbols, they as cellular entities are self-regulated and do perform actions. They have knowledge. They do not have meta-knowledge.

Jim Hamlyn said...

Bruce, I think you are neglecting our skills of nonverbal representation without which the acquisition of language would be impossible in the first place.

Do feral humans lack goals? Do infants? Do non-human animals. Do we need language to perceive too? Surely you don't think that perception lacks intention?

You say that paramecia are not conscious yet you claim that they have knowledge. Are you saying that I know how to digest my dinner? How is that knowledge in any meaningful sense? Can I share it?

If there is no way to share our knowledge — to represent it — then I don't think it qualifies as knowledge. I didn't learn to digest my dinner and nor can I forget how to do it. Both of these are characteristic features of knowledge.

Bruce said...

"Both of these are characteristic features of knowledge." Jim Hamlyn

Both of these are characteristic of knowing that you know, meta-knowledge. Our bodies have neurological and metabolic pathway "knowledge" - otherwise how could bodies self-regulate?

John S said...

Actions clearly imply agency and knowledge is an action. Therefore, all of you symbol pushers who deny agency are wrong. However, you guys are right about Materialism. Agency doesn't imply irreducible Mental Entities.

Jim Hamlyn said...

Bruce, I'm really scunnered by your responses. You now talk of how our bodies "decode" but you now surely mean some technical version of encoding that conceals untold commitments to further knowledge. How does the organism know how to decode its knowledge? At some important point even you have to assume a cutoff between sophisticated mechanisms and intelligent capacities otherwise you will need to defend knowledge all the way down to the molecular level. If you do that then you will still need to establish the basis of skill.

Perhaps we can focus on skill for a while. I hold that skills and knowledge are the same thing. What's your view? Would you want to distinguish between them?

Bruce said...

"How does the organism know how to decode its knowledge?" Jim

That is a strange question. Why should higher umwelt capacities be projected onto lower umwelts? You have heard of replication of DNA and protein synthesis in ribosomes, and of course natural selection.

Gottlob said...

Calm down people! Thank you Thomas Brandon for your support. By information giving a general law I did not mean the information forms laws but rather is a relation that is gathered from a law. Like say Boyle's law about the pressure and volume relationship of a gas at a constant temperature are inversely proportional, the information from this can be an innumerable number of things. It can be the logical relationship of the constant to the pressure, or the inverse proportionality of the pressure to volume, or the uniformity of the temperature. It maybe the pressure as a physical substrate, or the units of volume taken, or the nature of volume of a gas. Maybe in information theory we are trying to frame a certain packet of a physical or logical relationship or unit and give it a singular status and see how it interacts with each other and other units of information. How these units interact with each other can be logically or physically, but according to Wiki I think it is taken as a statistical measure. Also Jim Hamlyn's point is very valid; is intent essential in the elaboration of an information unit or not? But Matt Bush's point is also valid; that a datum is essentially a type of relationship or relata. This is I think the view of Frege and Russell.

Scott said...

Facially, per the OP, the Bush position seems absurd. I glanced at one of Ladyman's articles and it seemed tedious and didn't immediately suggest a clean and easily statable position.

Jim Hamlyn said...

Scott, going from previous discussions we have had you seem to have thought a lot about the necessary and sufficient conditions for intentional behaviours: actions. What's your take on knowledge? Does a bacterium have knowledge that it deliberately acts on in your view?

Bruce said...

Why in the world would you put "deliberately" in that question?

Jim Hamlyn said...

Because acts are intentional. You don't deliberately digest things but you do deliberately eat.

Digesting is a behaviour, eating is an act.

Bruce said...

This shows the sterility of word meaning analysis. What problem is this debate about knowledge trying to solve?

Jim Hamlyn said...

The debate is centred around the conceptual distinction between perceptually mediated behaviours (actions) and the many efficacious behaviours of sensory responsiveness. Check out Fred Adams' recent work, look up "blindsight" and think of the difference between what skills you can demonstrate when you are asleep in comparison with those you can demonstrate when you are awake. That should give you some idea of the crucial importance of this distinction.

Gottlob said...

Digesting is not a behaviour, it is a process. And eating is not a single act per se it is made up of a variety of acts like chewing, salivating, rolling your tongue, swallowing, etc.

Jim Hamlyn said...

Ok Gottlob, that's a fair challenge. What we want to do though is distinguish between processes like crystallisation say, and processes like digestion. I'm suggesting that sensorily mediated behaviours like digestion or iris dilation or galvanic skin response are not merely processes of the same sort as crystallisation because they are fundamentally responsive but not perceptual.

Gottlob said...

There's a word for that. Reflexive. As in a reflex arc. Or perhaps you are thinking of feedback mechanisms?

Gottlob said...

Eating, urinating, defecating, giving birth; these are I think neither acts nor behaviours, but lie somewhere in between. What is the proper word for it?

Jim Hamlyn said...

If they are perceptually mediated, then they are actions. If they are not perceptually mediated, then they are mere behaviouristic responses.

Derek Melser helpfully offers the term "natural processes" which accords well with your point. Nonetheless he also distinguishes between these and actions. http://www.derekmelser.org

Matt said...

John: for the record, I raise the scientism banner proudly. That said, the only thing that would scare me about abandoning scientism is the small risk of sounding like you

As for data, when you say "uninterpreted" I prefer "unmapped"; meaning we haven't figured out the structure in the data yet. "More research is needed in this area."

Heh, on data being relata, actually, data is a relation, and so is a detector, and a sender. There's no need for intrinsic objects or properties here, and in fact the labels only serve as signposts for scientists to orient themselves in the structure. Neat huh?

Scott said...

Jim Hamlyn, I agree with your analysis. I'm not sure what is driving this discussion. This seems to me an example of radical speculation and not true philosophical analysis. However, I'm sidelining on this one just hoping those against the motion can express a coherent meaningful position. So far no good

Jim Hamlyn said...

Thanks Scott, much appreciated.

Matt, More sneering. Really? Once again, what do you hope to achieve by such contemptuousness? You have made your position clear, why spoil it by marking your territory with barely veiled hostility?

Scott said...

Is Matt attempting to reduce the intentional to the nonintentional? Is consciousness being reduced to unconsciousness? Is agency being reduced to the causal powers of inert matter or lifeless matter?

Scott said...

Smacks of an ontological reduction

John S said...

Yes Matt Bush is reductionist. What he fails to realize is that even when formal systems correctly map aspects of our universe, they still require assent from the Scientific Community with empirical justification to amount to knowledge.

Gottlob said...

Jim, behaviouristic responses usually involve perception. And actions also involve intention and perception. Physiological processes on the other hand are completely mechanistic; they usually have the same properties as crystalisation, condensation, evaporation, distillation, etc. However physiological responses can be associated with complex behaviours (e.g flight or fight response, heat seeking in cold weather, hibernation, etc). Physiology and behavioural sciences are separate subjects, though there is much overlap.

Scott said...

I wonder how Matt handles normativity? I wonder if formal systems are subject to legal systems with sanctions for disobedience to norms. Can't wait to see how this reduction explains morality. Lol

John S said...

Hume said, "a wise man Proportions his beliefs to the evidence". It seems to me this is a true ethical judgment. Therefore, scientific knowledge presupposes ethics.

Jim Hamlyn said...

Gottlob, "Physiological processes" indeed yes, so that was the term you were looking for. Excellent, I'm happy with that. So we need -- or at least I am suggesting that we need -- to distinguish between physiological processes on the one hand and actions on the other. When we are asleep, under general anaesthetic or in a coma etc. there are only physiological processes going on and and no action. Most of our sensorily mediated behaviour is of this kind even during wakefulness. So I am claiming that action emerges with perception because perception is a skilful ability that comes with learning and knowledge both of which rely upon our capacity to demonstrate -- to represent our causal encounters -- to other similarly endowed perceivers.

John S said...

Yes I think you are on the right track Jim Hamlyn. I think we need a Theory of Human Action to account for knowledge. That is, Knowledge in the sense of "knowing that" rather than just "knowing how".

Jim Hamlyn said...

The "symbol pushers" won't like it but I would say that knowing-that is just another form of knowing-how. It's a knowledge of how to offer a symbolic token that will be accepted by other symbolic token users.

Bruce said...

Jim, The point you have made has been made passionately, but it is hardly momentous to acknowledge that language has emerged along with propensities to manipulate and exchange symbolic representations. I would caution however against a view of one to one correspondence between so-called beliefs in minds and the linguistic entities that are made and shared. It is the action of producing the linguistic entities that creates a new world and changed propensities. I engage in no debate against emergence, as I have stated earlier.

However one must also address the chicken and egg problem. What precedes knowledge? An earlier kind of knowledge. Perceptions are not primary.

“The problem 'Which comes first, the hypothesis (H) or the observation (O)?' is soluble; as is the problem, 'Which comes first, the hen (H) or the egg (O)?' . The reply to the latter is, 'An earlier kind of egg'; to the former, 'An earlier kind of hypothesis'. It is quite true that any particular hypothesis we choose will have been preceded by observations - the observations, for example, which it is designed to explain. But these observations, in their turn, presupposed the adoption of a frame of reference: a frame of expectations: a frame of theories. If they were significant, if they created a need for explanations and thus gave rise to the invention of a hypothesis, it was because they could not be explained within the old theoretical framework, the old horizon of expectations. There is no danger here of infinite regress. Going back to more and more primitive theories and myths we shall in the end find unconscious, inborn expectations.”
Karl Popper, Conjectures an Refutations (1963) p47.

John S said...

The symbol pushers forget that we are born with innate capacities. This is our starting point (not Locke's Blank Slate). After that we become infant Scientists learning from trial and error. Only much later do we join community's of other beings like us that modify our beliefs by criticism. Some of us join academic communities of other beings that share our specialized interests and modify our beliefs via academic journals. But what we cannot forget is that this whole process is driven by our human actions in the world.

Matt said...

Jim: that's a vague question and appropriate heuristic selection depends on whether we're using the manifest image and it's associated heuristics, or the scientific image and the information-theoretic scale (roughly equivalent to some special science - psychology, neuroscience, classical physics, etc) we are investigating.

«Oldest ‹Older   1 – 200 of 277   Newer› Newest»

Post a Comment