Interview with Bertie. Another one.

Reconstructing the Chinese Room (III)

Replies and objections*

The ‘System reply’

The reply: The person in the room is not a computer or software it is a CPU (“as it were”) running a software. So of course the person does not understand Chinese, but the whole system does.

Searle’s reply: (1) The system has no more means of attaching content to the symbols that the “CPU-person” does. Give him the walls, pen, paper etc. Conjunctions of that person with bits of paper and whatnot will not give cognitive features to non-cognitive entities, that is, a program (2) And even if it does, consider that the person internalizes everything by memorizing every single bit of what’s outside the Chinese room. (3) This reply would have the absurd consequence that “mind is everywhere.” For instance, “there is a level of description at which my stomach does information processing” there being “nothing to prevent [describers] from treating the input and output of my digestive organs as information if they so desire.

This in fact a very common reply to functionalism and it is based on the idea that “non-standard realizations” of cognitive phenomena will not be cognitive.

The ‘Robot reply’

The reply: There is no understanding involved because there aren’t right kind of causal relations between the symbols, the program and the referents. An appropriately programmed robot, in rich causal contact with its environment (capable of reacting to stimuli, memorize sense-date, moving etc.) would indeed have genuine understanding.

Searle’s reply: (1) First of all, this concedes the soundness of CRA. To insist that syntax plus external causation would produce semantics is to agree that it otherwise wouldn’t. (2) Intentionality is irreducible, that is to say, you cannot reduce it to non-subjective phenomena (such as the action of stimuli upon some receptors). (3) But give the man in the Room all the senses and still no understanding of Chinese would occur.

The ‘Brain Simulator reply’

The reply: What Searle rules out is the possibility that what the computer does is simulate the actual activity of the brain. Thus, classic Strong AI aside, an appropriately configured and trained connectionist network would have the genuine psychological properties and would do so “solely in virtue” of having the properties in question.

Searle’s reply: (1) “I thought the whole idea of strong AI is that we don’t need to know how the brain works to know how the mind works” (2) The computational power of a simulated neuronal network is – so the Church-Turing thesis says – the same as that of a Turing-machine. In other words, ‘any computation you can do on a parallel machine you can do on a classical machine” (3) The Chinese Gym: Imagine many-many-many human beings doing whatever a single individual firing of a neuron does inside a brain. Still, none of the individuals, nor the whole room would understand Chinese. (3) Imagine a man operating water-pipes instead of symbols. And let’s assume that the water pipes work, formally, in the exact same way as neurons do. Then he would be simulating a brain and still not have understanding of Chinese.

Preston (2002) identifies further objections which – although haven’t received a name or a tag – might be of equal importance as the ones above. I will note here their main points also because neither Damper (2006) nor the two internet articles from Stanford Encyclopedia of Philosophy & Internet Encyclopedia of Philosophy say anything about them. I’ll also have a go at some labels…

The “running program” reply

The reply: The deal with the Chinese room was that the man inside was implementing a program and that we make decisions about this man by regarding the properties of implementing (or running) a program. However, Searle’s premise is that Syntax is not sufficient for semantics – in other words that the program (not a program run) is not made up of anything else that syntax. Although this might be the case, the question still remains whether a running program is doing anything different. Hauser (2002) says – not very convincingly, if I may say – that it does, it must do, and recent empirical studies had shown how this could be the case. Whether one chooses to call this intentionality or not is a matter of terminological dispute.

The no-rule reply

The reply: Even the simplest rule-following operations require agents capable of exhibiting the capacities characteristic of normativity: they must understand the rules being followed, be capable of explaining, justifying, and correcting what they (and others) do by reference to the rules in question. Both Turing and his followers mis-characterized his achievement. What they had done was to identify the set of functions which could be processed without intelligence. In this view, what weak & strong AI is doing is in fact showing how ultimately mechanical objects can appear to be performing what humans are actually performing. They only appear to be doing it because they don’t do it in the same way and by using the same means.

______

*The series “Reconstructing the Chinese Room” follows some of the articles in:

Preston, J. & Bishop, M. (2002). Views into the Chinese Room: New essays on Searle and artificial intelligence. Oxford: Clarendon Press.

The first three posts will follow:

Preston, J. (2002). Introduction. In Preston, J & Bishop, M. Views into the Chinese Room: New essays on Searle and artificial intelligence, pp. 1-51. Oxford: Clarendon Press

Other works cited here are:

Damper, R. (2006). The logic of John Searle’s Chinese room argument. Mind, 16, 163-183
Hauser, L. (2002). Nixin’ Goes to China. In Preston, J & Bishop, M. Views into the Chinese Room: New essays on Searle and artificial intelligence, pp. 123-144. Oxford: Clarendon Press

Reconstructing the Chinese Room (II)

An informal look at the argument*

The actual text of the CRA (Chinese Room argument) could be found anywhere on the Internet where the subject is tackled. I know I should not add one more to it, but here it is. This version is a revised one from his (1984) Reith lectures Searle held live on BBC:

Imagine that a bunch of computer programmers have written a program that will enable a computer to simulate the understanding of Chinese. So, for example, if the computer is given a question in Chinese, it will match the question against its memory, or data base, and produce appropriate answers to the questions in Chinese. Suppose for the sake of argument that the computer's answers are as good as those of a native Chinese speaker. Now then, does the computer, on the basis of this, understand Chinese, does it literally understand Chinese, in the way that Chinese speakers understand Chinese? Well, imagine that you are locked in a room, and in this room are several baskets full of Chinese symbols. Imagine that you (like me) do not understand a word of Chinese, but that you are given a rule book in English for manipulating these Chinese symbols. […] Now the point of the story is simply this: by virtue of implementing a formal computer program from the point of view of an outside observer, you behave exactly as if you understood Chinese, but all the same you don't understand a word of Chinese. But if going through the appropriate computer program for understanding Chinese is not enough to give you an understanding of Chinese, then it is not enough to give any other digital computer an understanding of Chinese. (Searle, 1984, pp. 31-32)

In the original one, an elaboration appeared with the same man receiving cards with signs in English – aka letters – but the purpose of this elaboration was merely to make the differences clearer. The structure of the argument was in every respect the same.

So what is the CRA about? Preston’s (2002) “Introduction” seems very sharp on this point. Here is his informal summary:

The central claims of Searle’s original paper are clear. Something is a digital computer in virtue of performing computations. But computations alone cannot, in principle, give rise to genuine cognition. Computation being nothing but the manipulation of symbols in accordance with purely formal or syntactical rules, something that is only computing cannot be said to have access to or know or understand the ‘content’, the semantic properties of the symbols it happens to be manipulating. […] But it is a conceptual or logical truth that syntax is not sufficient for semantics. Computer, therefore, cannot be credited with understanding. (p. 19)

What I see in his commentary is a very subtle glide between two fundamentally different interpretations. Since they are different, and since this is not made explicit but appears on and off in Preston’s contribution, I think the central claims of Searle’s original paper are not in the least “clear”. Informally, we could say that CRA might be putting forward two different claims (either/or and maybe even both):

(1) While something digital can simulate intelligent performance, nothing of the sort can duplicate it

(2) While something digital can simulate intelligent performance, nothing of the sort can be said to have duplicated it.

Notice also some of Preston’s choices: “can be credited with understanding” (as opposed to “understands”, simpliciter) or “cannot be said to have access to…” (as opposed to “have access to”, simpliciter).

Interpretations and implications

There has been a lot of interest in distinguishing different types of replies to CRA. Some, however, are more noteworthy than others. It has often been stressed that ‘understanding a language’ is not a relevant component of a Turing test if the computer is not “packed” with a language-learning program. But, of course, this would only mean that Searle is to construct an analogous scenario where the task in question is learning Chinese not understanding Chinese.

Also, some replies have focused on the difference between “if squiggle-squiggle, then squaggle-squaggle” and (a) the actual formalisms in computer programs and anyway (b) the actual pragmatics of language. To both, however, there is an easy rebuttal: (a) although they become more and more complex, computer programs become thus by adding up a collection of such rules which run in virtue of what squaggle-squaggle is related to that squiggle-squiggle; (b) of course it is not, but this is an assumption that both parties agree with on the debate, namely, they both assume that computer programmers will be able to replicate language in that way [In other words, AI says, If …, then yes …, while Searle says, Even if…, then it would still not … - but both are ready to assume the if-part.]

Each of the more serious replies will be spelled out in later sections.

What is curious is that even though Searle put forward a reconstruction of his own of what CRA is intended to establish as an argument, critics have rejected that very reconstruction. The 3 premise Searle himself ascribes to his own thought-experiment are the following:

(1a) Programs are purely formal

(2a) Minds have mental content

(3a) Form is never the same as or sufficient for semantics

Therefore,

(4a) Programs are not sufficient for minds

This reconstruction is, I think, deceivingly simplistic – maybe part of the reason why some have dubbed it ‘the Brutally Simple Argument’. For one thing, the elements in the Chinese Room (the man, the squiggle, squiggle, the deceiving output etc.) do not appear. How is that then a reconstruction of how Searle set his argument if almost the whole of is argument is not present.

Here is the reconstruction Preston chooses to put forward. It is not utterly different from Searle’s but it at least brings into the reconstruction elements from the thought experiment itself. (Still, one might wonder why it is not the same as Searle’s):

(1b) The person in the room has access only to the formal features

(2b) To understand Chinese, the person would need access to semantic features

(3b) No set of formal features is sufficient for understanding

None of the reconstructions above seems to be very loyal to the actual text and in any case none are justified in any way. They’re just put forward either by Searle or by his Critics.

_______

*The series “Reconstructing the Chinese Room” follows some of the articles in:

Preston, J. & Bishop, M. (2002). Views into the Chinese Room: New essays on Searle and artificial intelligence. Oxford: Clarendon Press.

The first three posts will follow:

Preston, J. (2002). Introduction. In Preston, J & Bishop, M. Views into the Chinese Room: New essays on Searle and artificial intelligence, pp. 1-51. Oxford: Clarendon Press

Reconstructing the Chinese Room (I)

Historical background*

1936

Alan Turing publishes a paper entitled “On computable numbers, with an application to the Entscheidungsproblem”. The problem referred to in the title is one posed by David Hilbert (don’t know if the problem is from those legendary 100 posed at the international colloquium…): is there any effective (i.e. mechanical & finite) way of determining, of any given statement in a formal system, whether or not it is provable in that system? Because he had set out to show that there is no such thing, i.e. that mathematics is undecidable, Alan Turing needed a precise definition of what “effective & finite” meant in this context. Turing came up with this: there is an “effective & finite” way of calculating the values of a mathematical function if it can be processed by a ‘computer’ – at that time the meaning of the term was that of ‘a human who computes’ – which he called the Turing machine. This was the first description of a very basic computer and its computer program.

Along with the description of this machine came the thesis (later dubbed ‘Turing’s theorem’) that the outputs of any one of these machines can be written as inputs to other Turing machines ad infinitum. Although it is impossible to construct and/or operate with such machines, the idea opened up an inspiration for those who believed the prototype of unlimited storage and perfect reliability.

But even assuming some functions are undecidable, what is this sub-set of decidable functions and how is it to be identified? The American logician Alonzo Church had posited in late 1933’s the idea that this set could be identified with the notion of recursive functions (a concept mathematically precise enough since it relied on the (by then) well known concept of lambda-calculus).

The Church-Turing thesis was born out of the unification of these two views. The set of recursive definitions, those functions which are computable by Turing machines, is the same as the set of functions which can be computed by any effective, mechanical, finite etc. method.

1950

Turing publishes “Computing Machinery and Intelligence” in which he recasts the can-computers-think-question into the form of a game: we can speak of it as intelligent if it renders impossible for a human “interrogator” to tell whether the output (coming from it) is that of a machine or a human mind.

Many attacked the psychological assumptions behind this test, the Turing test. [Rather preoccupied about gluing one’s surname to each of one’s productions, isn’t he? I wouldn’t be surprised if I found out colleagues started calling him Turing’s body…] It was accused of being too behavioristic – that is, assuming that the meaning of the statements about one’s psychological states can be reduced to statements about his bodily motion. It was also accused of being too operational – the epistemological view that the concepts of science (“intelligence”, here) should be defined “in terms of” the number of operations we need to apply in order to verify them

1960s

The idea of a Turing test, together with the criticisms it received, was taken over by the American philosopher Hilary Putnam who stated what came to be known as the functionalist view of the mind. According to this view, mentality is a matter of functioning not of substance. In other words, it doesn’t matter what the mind is made of, what matters is what it does, how it works and what is capacities are. By “to matter”, here, Putnam meant mostly “to be able to speak meaningfully of”. Mental concepts, in other words, are functional concepts.

But then, if this is the case, then whatever simulates mental phenomena amounts to duplicating them – since all that there is to mental phenomena is their functioning. There’s no reason, Putnam implies, not to credit computers running appropriate programs with mentality. In fact, both minds and computer can be described in two ways: by referring to their hardware (this impulse passed from there to there, thereby enabling this and that and that causing such-and-such which gave a certain output) or to their software (the computer calculated 2+2). Hardware/software – brain/mind: the analogy worked perfectly.

Of course, many contested this view as well. It was primarily the serial workings of the Turing test that troubled scientists. We know that the mind is capable of more than one state at one time (at least a state and its introspective duplicate). A Turing machine, however, can only be in one state at a time – even if its program can process a collection of series.

Another, more profound, criticism to this idea was the point that psychological states are incompletely described if one doesn’t take into consideration their causal connection with sensory or other behavioral outputs.

Artificial Intelligence: Strong & Weak

Starting mid-1950s, but producing observable results only some ten years later, the artificial intelligence research programme, if one can speak of such a unitary entity, started by looking into the connections between neurons and symbols. First concerned with the classic task of computing, AI researcher were gradually interested in games (chess, guess-games) and then further in what might be called “understanding” (language, planning, learning etc.)

According to the view they held as to the aim of this programme, two versions of AI have been distinguished. The Weak AI is committed merely understanding human psychological phenomena and trying to replicate these so that a computer can produce the same results without the implication that such a computer would thereby have the psychological states. The more radical view, known as Strong AI, is committed to the idea that such a program would genuinely have those psychological states (remember “functionalism”?). As Searle later put it, that such a program would “have a mind in the same sense you and I have”.

[What seems apparent to me is that these “versions” of AI are different neither in their empirical commitments nor in their theoretical scope. They are methodologically different; that is, they are different in what they take as good or bad method. For one, it is cool to say that the super-computer is having this or that psychological state. For the other, it is not. I also think that the methodological import of these versions can be easily omitted. One of Preston (2002) quotes is an example of how this can be done:

Searle, on the other hand, claims to have found an argument that undercuts the idea that electronic digital computer can be said to exhibit any of the contested psychological capacities purely in virtue of their programs. Philosophers certainly have no insight into what technical tasks programmed machines might be able to perform, or when. But they can have a say about how it makes sense to characterize the abilities in question. (p. 16)

So far, from the italics above, it is clear (I think) that the bone of contention is methodological. Whether concerned with the property of terms such as “psychological states” or with how is neat and profitable to act as a scientist, they are prescriptive standpoints. Preston, however, seems to conclude the paragraph by judging CRA with having made a descriptive point about what can and cannot be the case (as opposed to “be said to be the case”). He continues:

Like anyone else they may, by using thought experiments for example, establish or refute thesis about what is logically possible.

______

*The series “Reconstructing the Chinese Room” follows some of the articles in:

Preston, J. & Bishop, M. (2002). Views into the Chinese Room: New essays on Searle and artificial intelligence. Oxford: Clarendon Press.

The first three posts will follow:

Preston, J. (2002). Introduction. In Preston, J & Bishop, M. Views into the Chinese Room: New essays on Searle and artificial intelligence, pp. 1-51. Oxford: Clarendon Press

An answer to the gap problem (?)

Brown, J. (2011). Thought experiments, intuitions and philosophical evidence. Dialectica, 65(4), 493-516.

The paper starts as follows: “What is the nature of evidence provided by thought experiments in philosophy?” What is surprising, then, in view of this undertaking, is that none of the towering literature on this matter is quoted or used. Unfortunately, instead of providing a fresh approach on this old chestnut, the conclusions supported in the paper are rather unsophisticated.

Brown identifies the problem as follows:

the psychological proposition that it seems to me as if the Gettier subject has a non-knowledge justified true belief is not in itself a counterexample to the JTB theory of knowledge. A genuine counterexample would be a case of justified true belief which is not knowledge (Brown, 2011, p. 500)

So yes, the old intuition-pumps criticism put forward by Daniel Dennett. If thought experiments are based merely on “what one would say” or “what one would feel” or “what one would accept” in the depicted scenario, how can we use thought experiments to support those solid, difficult, objective theories? Using Brown’s terminology, and putting it a bit crudely, how can psychological facts support non-psychological propositions? The assumption underpinning the relevance of this question is that there is a gap (of the subjective/objective kind).

there is a gap between one’s perceptual evidence and what it is taken to be evidence for, in the sense that the proposition that one is having an experience as of p does not entail that p is the case.

Brown diligently considers many views and viewpoints on this matter and some other ones more or less related. She spends a lot of time with the similar problem of perception (how can perception justify theoretical knowledge? etc.)

Finally, her answer is this. I’m quoting it because I don’t understand it.

one could close this apparent gap by appeal to an externalist approach to justification and/or knowledge. Suppose that whenever one has such an experience, one forms an appropriately related belief. For instance, that when one has an experience as of a large barking dog in front of one, one forms the belief that there is a large barking dog in front of one. On an externalist approach to justification, such as reliabilism, as long as the appropriate external relations hold, the beliefs so formed are justified.

One might hope to apply this solution to the gap problem in the case of perception to the gap problem facing the psychological view of thought-experiment evidence. Suppose that, in fact, the method of forming beliefs about the nonpsychological subject matter of philosophy on the basis of the relevant psychological propositions is reliable. Combining this supposition with a reliabilist approach to justification has the result that beliefs formed in this manner are justified (p. 513)

I honestly do not see what this “combination” means. I understand she is saying this: “Well, remember how subjective propositions can support subjective propositions? Take that, and apply it to our situation, where subjective propositions must support objective ones and say that the appropriate external relations must hold”. But how is this an answer to the problem? One should immediately respond: “Well sure, there must be some extra conditions applying to this step since we know that sometimes subjective/psychological data is misleading and leads to unjustified objective/non-psychological data. We know that. We know that not all of those inferences are justified and we know that the ones that are justified are as such in virtue of the criteria. But the question here concerns precisely these criteria. It is no news nor solution to say – ‘Well, you see, some criteria must hold, criteria concerining the passing from the subjective to the objective’”.

All in all, I guess, this boils down to answering the question of expertise: When is appeal to expertise to be trusted? Because, in thought-experiments that appeal to intuitions – if there are such things, and I think there are – it is the philosopher’s claim to his expertise on one’s intuition that functions as argument.

Unfortunately, it might turn out that no universal answer is possible and that the criteria are highly field-dependent.

A TE against JTB

The original Gettier-problem

In a three-page paper titled Is Justified True Belief Knowledge? (1961), Edmund L. Gettier claimed to have struck at the very foundation of a then rather firm tradition in epistemology, a tradition which considered knowledge as justified true belief (JTB). What is interesting about this paper is that its method is “thought experimental” – if such an adjective may be used.

Since its publication, many have acknowledge the fair death of JTB and many have tried a rebuttal of the two thought experiments it contains (as it seems quite often to be happening with thought experiments). There are standard replies and standard replies to replies, a very good exposition of which can be found in the Wikipedia article on it. Let us first take a brief look at the article.

“Is justified true belief knowledge?”

Gettier first identifies an established account which defines knowledge by reference to three sufficient conditions:

S knows p iff (a) p is true, (b) S believes that p, (c) S is justified in believing that p

The two philosophers pointed towards are R. Chisolom & A. J. Ayer. To this move some have responded as follows:

According to the inherited lore of the epistemological tribe, the JTB [justified true belief] account enjoyed the states of epistemological orthodoxy until 1963, when it was shattered by Edmund Gettier... Of course there is an interesting historical irony here: it isn't easy to find many really explicit statements of a JTB analysis of knowledge prior to Gettier. It is almost as if a distinguished critic created a tradition in the very act of destroying it. (Plantinga, 1992)

This, however, is a problem epistemologists and maybe historians of philosophy must deal with. As far as the thought experiments are concerned, even if one philosopher advanced a theory of the sort – and the theory itself is not total bogus – then an inquiry into the debate itself is not a useless assignment.

Before the definition of “S knows p if (a) (b) and (c) hold”, Gettier introduces two priciples which will be used in the thought experiment.

(i) It is possible for S to be justified in believing a proposition p “that is in fact false”

(ii) If S is justified in believing p and p "→" q, then S is justified in believing q if S is the one making the deduction and if S is making the deduction from p.

Again, one might feel the need to claim that these principles are not immediately acceptable within the debate. I think a similar criticism applies to both. First, the subtle insertion of “in fact” in (i) is problematic. If p is “in fact” false, then S is “in fact” not justified in believing p – but only “apparently” or “as far as he knows”. It’s hard to tell whether this affects the thought experiments and their conclusion, but nevertheless a preliminary question could have been addressed: Does S know p if p is true (according to him) but false (according to us)?. Second, the possible ambiguity of “justified” appears in (ii): S might think he is justified in believing p and be really justified in believing in q. Again, I don’t know if these points affect the thought experiment, but I don’t think they’re completely irrelevant either.

John & Smith & coins & jobs

Suppose that Smith and Jones have applied for a certain job. And suppose that Smith has strong evidence for the following conjunctive proposition:

(d) Jones is the man who will get the job, and Jones has ten coins in his pocket.

Smith's evidence for (d) might be that the president of the company assured him that Jones would in the end be selected, and that he, Smith, had counted the coins in Jones's pocket ten minutes ago.

Proposition (d) entails:

(e) The man who will get the job has ten coins in his pocket.

Let us suppose that Smith sees the entailment from (d) to (e), and accepts (e) on the grounds of (d), for which he has strong evidence. In this case, Smith is clearly justified in believing that (e) is true. But imagine, further, that unknown to Smith, he himself, not Jones, will get the job. And, also, unknown to Smith, he himself has ten coins in his pocket. Proposition (e) is then true, though proposition (d), from which Smith inferred (e), is false. In our example, then, all of the following are true: (i) (e) is true, (ii) Smith believes that (e) is true, and (iii) Smith is justified in believing that (e) is true. But it is equally clear that Smith does not know that (e) is true; for (e) is true in virtue of the number of coins in Smith's pocket, while Smith does not know how many coins are in Smith's pocket, and bases his belief in (e) on a count of the coins in Jones's pocket, whom he falsely believes to be the man who will get the job.

(Before we move to the second one, let us notice the similarity between this thought experiment and a classic scenario depicted by K. Donnellan in his paper on Reference and definite descriptions (1966). In that paper, Donnellan was arguing against the classic positions on reference assumed up to that point by Russel and Strawson by conjecturing a distinction between two uses of definite descriptions, a referential and an attributive one:

To illustrate this distinction, in the case of a single sentence,consider the sentence, "Smith's murderer is insane." Suppose first that we come upon poor Smith foully murdered. From the brutal manner of the killing and the fact that Smith was the most lovable person in the world, we might exclaim, "Smith's murderer is insane." I will assume, to make it a simpler case, that in a quite ordinary sense we do not know who murdered Smith (though this is not in the end essential to the case). This, I shall say, is an attributive use of the definite description.

The contrast with such a use of the sentence is one of those situations in which we expect and intend our audience to realize whom we have in mind when we speak of Smith's murderer and, most importantly, to know that it is this person about whom we are going to say something. For example, suppose that Jones has been charged with Smith's murder and has been placed on trial. Imagine that there is a discussion of Jones's odd behavior at his trial. We might sum up our impression of his behavior by saying, "Smith's murderer is insane." If someone asks to whom we are referring, by using this description, the answer here is "Jones." This, I shall say, is a referential use of the definite description (Donnellan, 1966, pp. 285-286)

This distinction, as well as the two scenarios, were afterwards (again) fiercely discussed by Kripke, Searle and Bach. I have posted a few times on this problem and its tradition, see here.)

John & Smith & Brown & cars

Let us suppose that Smith has strong evidence for the following proposition:

(f) Jones owns a Ford.

Smith's evidence might be that Jones has at all times in the past within Smith's memory owned a car, and always a Ford, and that Jones has just offered Smith a ride while driving a Ford. Let us imagine, now, that Smith has another friend, Brown, of whose whereabouts he is totally ignorant. Smith selects three place-names quite at random, and constructs the following three propositions:

(g) Either Jones owns a Ford, or Brown is in Boston;

(h) Either Jones owns a Ford, or Brown is in Barcelona;

(i) Either Jones owns a Ford, or Brown is in Brest-Litovsk.

Each of these propositions is entailed by (f). Imagine that Smith realizes the entailment of each of these propositions he has constructed by (f), and proceeds to accept (g), (h), and (i) on the basis of (f). Smith has correctly inferred (g), (h), and (i) from a proposition for which he has strong evidence. Smith is therefore completely justified in believing each of these three propositions. Smith, of course, has no idea where Brown is.

But imagine now that two further conditions hold. First, Jones does not own a Ford, but is at present driving a rented car. And secondly, by the sheerest coincidence, and entirely unknown to Smith, the place mentioned in proposition (h) happens really to be the place where Brown is. If these two conditions hold then Smith does not know that (h) is true, even though (i) (h) is true, (ii) Smith does believe that (h) is true, and (iii) Smith is justified in believing that (h) is true

This one is a bit simpler since it does not involve any definite descriptions. However, the objections to (ii) do apply: it is not clear in what sense S is justified in believing (h) if Jones does not own a Ford. Some early responses attempted to save JTB by rejecting (ii) – thus, by adding the condition that the belief not be inferred from false premises. An example of this is given by Wikipedia:

After arranging to meet with Mark for help with homework, Luke arrives at the appointed time and place. Walking into Mark's office Luke clearly sees Mark at his desk; Luke immediately forms the belief 'Mark is in the room. He can help me with my logic homework'. Luke is justified in his belief; he clearly sees Mark at his desk. In fact, it's not Mark that Luke saw; it was a marvelous hologram, perfect in every respect, giving the appearance of Mark diligently grading papers at his desk. Nevertheless, Mark is in the room; he is crouched under his desk reading Frege. Luke's belief that Mark is in the room is true (he is in the room, under his desk) and justified (Mark's hologram is giving the appearance of Mark hard at work).

A further objection might be advanced, this time as to what Gettier means by having justification for something. One might interpret the phrase as "there being something out there that can justify one's belief". One might choose to say Jones has justification for believing that man landed on the moon when Jones believes that unjustifiably, i.e. without considering the justificatory evidence. But one might also interpret it as "there being something out there that can justify one's belief and that person using it as such". In this case, we could not say the same about the situation with Jones' belief of man landing on the moon; there is a justification out there, but Jones does not have it. 

Gettier did not propose any solution to this problem, which is how he managed to keep his paper so short.

PS: In order for the discussion not to be jumbled one might formulate the following question: How can two discussants agree on a standpoint but disagree on the arguments which must be taken as a good defence of that standpoint? At first glance, it is simple: they just don’t agree on the arguments but they (happen to) agree on the conclusion. But is this even possible? They might also disagree on the standpoint “P because Q”, or something similar.  

Alvin Plantinga (1992). Warrant: The Current Debate. Oxford University Press. pp. 6–7

The world & everything

Leibniz’s Monadology (1714) is anything but a cosy, Sunday-afternoon read. We know Bertrand Russell was an avid reader of Leibniz’s work and its exegesis. Russell even wrote a book on it, A Critical Exposition of the Philosophy of Leibniz (1900). We know that. We know that and we cannot understand why. Leibniz's writing is callous and it is frustrating.

Nevertheless, here and there we find a piece of argumentation. And argumentation in those days was a hot bowl full of spaghetti with more geometrico, thank you. For instance, in order to prove that the world is made out of teeny-tiny souls – “monads” – which receive unique and infinite understanding from God, he went as follows.

First, two postulates:

I. Every substance is either simple or compound.
II. A compound substance is composed of simple substances.

Nice. But let one not be fooled by simplicity. These be starting points treacherous and deluding! Anyway, a conclusion certainly follows.

III. Every substance is either simple or composed of simple substances.

We then lay down two additional postulates:

IV. Each material substance has a divisible extension.

Has a what? An extension of a substance is, in Descartes’ terminology, is the amount of space matter occupies. Let it be said that, by borrowing this scholastic concept, Descartes was perfectly mudded in an old medieval metaphysics as to the distinction between mind and matter (intension and extension). But be it. The next postulate is:

V. Nothing that has a divisible extension is a simple substance.

Exactly, right? Because it is a divisible extension. But take III, IV & V, and state it clearly so that Russell can hear you:

VI. No simple substance is a material substance.

And, if not material, we know what’s left:

VII. Every substance is either material or spiritual

Hence,

VIII. Each simple substance is spiritual.

Good night!
PS: For those sufficiently sadistisch, check the Monadology for why every monad is a perceiving being!

A dilemma for thought experimenters

Thought experiments do not have a life of their own ... because of the following dilemma. We either have experience with a situation that is relevant to the thought experimental situation or we don’t. If we do, then our thought experiment just tells us what we know from concrete experiments (or observations). But if [we] don’t, then the thought experimental result amounts to mere guesswork.

Sure, but it depends on what you mean by "guesswork". In a sense, if it means "not being based on direct or reported experience", then yes. Now, many other forms of whatever you might think of as evidence (arguments, for instance) will be guesswork just as much. If, however, "guesswork" means "random conclusions that draw upon intuition" then sure again, but then we're agreeing to a rather different conclusion, for the suggestion is that the outcome is irrational. One way out of this dilemma was famously supported by Norton (1996), who recognizes the split in the road but saw a way out.

Norton, J. (1996). Are thought experiments just what you always thought? Canadian Journal of Philosophy, 26(3), 333–366.

Reiss, J. (2002). Causal inference in the abstract or seven myths about thought experiments. Technical Report CTR 03/02, Centre for Philosophy of Natural and Social Science, London School of Economics, London, UK

The transactional/interactional distinction

This distinction is often made in connection with two purposes (or functions) of language use. To use language is to attempt to achieve both at some level, with more or less interpretation needed to discern or explain these functions. Similar distinctions such as representative/expressive (Buhler), referential/emotive (Jakobson), and ideational/interpersonal (Halliday) follow a similar path.

The first one then is the transactional function: we (i.e. humans) use language to send messages with content, to send, that is, a representation of a non-linguistic content into a linguistic form. In the history of linguistics nobody was narrow-minded enough to claim that this is the only function of language, but often analysts have abstracted away from other functions in order to study what is “delivered” from one party to the other. The term “propositional content” is often used by semanticists who wish to state in very precise terms what a proposition communicates, regardless, as it were, from where and why and by whom it is uttered. It is assumed, as it where, that what the speaker primarily has in mind is a transfer of information.

The second function is the interactional function: we (i.e. humans) use language to establish and maintain various sorts of social relationships. Sociologists and anthropologists often speak of the phatic function of language and are keen to point that some of our everyday talk can even be said to have this function above the first one. As Brown & Yule (1983) notice that

“when two strangers are standing shivering at a bus stop in an icy wind and one turns to the other and says ‘My goodness, it’s cold’ it is difficult to suppose that the primary intention of the speaker is to convey information. It seems much more reasonable to suggest that the speaker is indicating a readiness to be friendly.”

Context & discourse analysis, love & marriage

 

To say that context influences communication is close to saying that the sport you are practicing influences what you might do during training, at competitions, what you must do for a medal and so on. Naturally however, while the influence of the context is recognized, theorists are not at all ready to jump in the same bucket of consensus when it comes to answering how this influence manifests itself and is to be studied.

To begin with, the question what is context? is itself quite spiny. I believe most natural language users would understand the notion in opposition to the text (it is the thing in which the text comes, which is different from the text in that it is somehow in the background). In a more technical language, the focal event - the actual text, if there is such a thing - is differentiated from the context. As a first, negative definition, it works just fine. Now, assuming there are instances of actual texts which somehow “do not need” the context – where does this context-thing first appear on a gradual scale of, let’s say, context-independence to high-context-dependence?

Most pragmaticians would start with deixis. That, at least, is something of a consensus. When I say “I said I will” I say something quite different than when you say “I said I will”. [Often, other factors such as tu/vous distinction (je/u in Dutch) have been called “social deixis”]. But what next? You can feel the ground getting shaky. One of Sacks’ mini-stories (The baby cried. The mommy picked it up) is dauntingly simple if it proves that huge chunks of socio-cultural contexts come into play at every step – for instance, the idea that mothers usually take care of their babies, that picking a baby up can be a response to a cry. Because, you might not have noticed, but it in the picked it up could very well refer to a bowl of soup: so how come it [sic!]… does not?

And it gets even worse in two ways. Context is not only highly resistant to theory-formation. It is also, first, something actors “negotiate” during speech. The focal event is built (see Gumperz’s contextualization processes) partly as presupposing but partly as generating the context. In at least one sense we can be sure of this statement: new text creates, in its bare minimum, new co-text for what preceded and what follows. Only, of course, considerations should land on the more abstract (social, institutional, culture) levels of text-context interaction.

Second, a distinction between several contexts for seems at once highly de-stabilizing and highly needed. When two parties are speaking and only those two studied, great. But there are social events with audiences and where the speaker and the communicator differ, and situation where the hearer and the receiver of the message differ, and finally situations where the audience and the over-hearers differ. In other words, it is not only a matter of systematizing (or ‘formalizing’) contexts and making their relationship with texts apparent, but also, in a way, ‘assigning’ them the party one wishes to study. Can anything be more difficult?

The (real) purpose of language…

A common explanatory attempt has been to point out the relationship between brain size and group size: the larger the groups in which animals live, the more there is a need for organization, the maintenance of complex and/or multiple social relationships, and often a division of labour, too. To interpret all the relevant information, more brain power is needed, which, in turn. stimulates the development of more complex communication systems. This process of the mutual reinforcement of strictly distinct properties within the same species might be seen as a form of 'co-evolution'. In conjunction with such observations the argument is often made that language developed primarily for bonding purposes, as an extension of primate grooming behaviour. Evidence for this is sought in the prevalence of social talk in ordinary conversation. (Verschueren & Brisard, 2009)

Thus,

Pragmatics as domain and pragmatics as perspective

 

Pragmatics does not study language as a system of signs, but as a tool. In French, we would say pragmatics deals with le langage not with la langue. Sometimes, this is explained by saying that pragmatic theories seek to explain user-choices. Everyone makes choices while using a language: one pronounces differently, orders words differently, means things differently, interprets things differently etc. The job of pragmatics is to explain these choices – not the reasons behind them, be it psychological, social etc., but the purpose with which they are made. (Notice, here, the relationship with the philosophy of pragmatism which was, broadly speaking, interested in the way processes, meaning-generating processes in particular, were connected to their purpose. The term, going even further back in time, is connected with Kant’s pragmatisch).

The cradle of pragmatic inquiry, whether in hindsight worthy of the current meaning associated with the term “pragmatics” or not, is usually situated in the almost Cartesian project of Morris (1938). Morris’ aim was to come up with a unified theory of signs (semiotics) which would bring together logicians, linguists, philosophers, psychologists, anthropologists, sociologists and others. Morris identified the three actors involved in semiosis and called them “sign (vehicle)”, “designatum” & “interpreter” (the labels are fairly intuitive). Semiosis is then a triadic relation, a correlation of the sign, the designatum and the interpreter which we might call, today, communication. Aside from this observation, six dyadic relations can be abstracted for further inquiry: sign-interpreter, designatum-vehicle etc. Pragmatics is then the relationship between signs and interpreters.

This definition has to be placed in the intellectual context of the emergence of semiotics as a philosophical reflection on the 'meaning' of symbols. often triggered by the use of symbols in science and hence related to developments in the philosophy or theory of science but soon expanded to all other domains of activity involving what Cassirer calls 'symbolical animals', i.e humans. (Verschueren, 2009, p. 3)

Inevitably, pragmatics came out eventually as a tremendously complicated and interdisciplinary endeavour, its roots being connected certain developments in psychology (Mead, Malinowski etc.), philosophy of language (Wittgenstein, Austin, Searle) and social theory (Hambermas). An “anthropological trend” also became visible early on, and here the works of Gumperz & Hymes (1972) Direction in sociolinguistics, Hymes’ (1974) Foundation of sociolinguistics, and Gumperz (1982) Discourse strategies stand as building-bloks. The anthropological approach was driven by the attempt to study discourse in context, without abstracting away from the social, cultural & “institutional” aspects of communication. It is why this approach is sometimes labelled as ethnographical, thus covering the field of conversation analysis (Garfinkel, Sacks, Schegloff etc.) The intuitive notion of context became more and more theory-laden.

The history of these developments can also receive an organization into schools, although many developments would be left aside if this approach were undertaken. The French school of linguistics (Ducrot, Anscombre etc.), the Prague school (Mathesius, Firbas etc.) are some of the more stable approaches.

A more common perspective is that of topics. It is sometimes acknowledged that pragmatics delineated its field as a result of developments in other fields. Some topics, as it were, ended up as constituting the subject-matter of a pragmatic endeavour and this is often referred to as “the wastebasket view of pragmatics”. For instance, the Chomskian distinction between competence & performance, for instance, simply leaves all subject connected to performance “outside grammar & semantics” – thus creating a new subfield as a by-product. (The twofold form vs. use distinction is today very much left as a stepping stone worthy of no more than historical interest). In the wastebasket of pragmatics we find highly interesting yet fairly disconnected subjects or topics (deixis, implicature, presupposition, conversation etc.). Although attempts have been made to unify this set of distinct approaches to distinct issues – Sperber & Wilson’s Relevance theory, Levinson’s (2000) Presumptive meanings – no theory has managed to gain consensus. Morris’ ideal of a unified theory of signs seems more and more an impossible undertaking.

Topical and methodological unity are required criteria if we are to speak of the field of pragmatics, complementary to other fields within the broad domain of linguistics. If there is methodology in logic and grammar – why shouldn’t it be in pragmatics?

Gazdar’s (1979) textbook famously introduced the idea of pragmatics as “meaning minus truth conditions” (alluding to Kempson’s (1975) theory of semantics as the study of truth-conditions) is but one path. The what-is-conventional vs. what-is-non-conventional approach (more or less accepted by Levinson’s (1983) textbook) is another one. Reference to context – i.e. defining pragmatics as the study of meaning in context – is yet another one. Verschueren (2009) observes in these approaches, with some degree of generality, a certain “fear of trespassing into the realm of sociolinguistics and psycholinguistics” (p. 12).

It should be clearly noted that multi- or inter-disciplinarily has not always been seen as an imperfection. Some textbooks even underline the benefits of such a state within a field of inquiry - for instance, one benefit would be the rise of well-structured theories on specific topics (Sperber & Wilson’s 1986 Relevance theory being one example). And, as some have pointed out, so long as agreement on what constitutes proof in pragmatics is not achieved, a grand project might even be to incoherent to be safe. It is then often asked: What counts as empirical evidence in pragmatics?

One way out of this is to refuse the view of pragmatics as a domain (with specific principles of rational and empirical justification) but as a perspective. A perspective on what? On whatever linguists (phonologists, semanticists, etc.) deal with. Some subjects from these fields will lend themselves to being studied from the perspective of pragmatics, some will not. This demarcation does not spoil the unity of the fields/domains/subject-matters in question. Verschueren (2009, p. 16) supports this tactic as follows:

There is at least one essential difference between pragmatics and what we have referred to as components of linguistics. In contrast with phonology with phonemes as basic units of analysis, morphology with morphemes, syntax with sentences, and semantics with propositions or lexical items, pragmatics cannot - without undue oversimplification – be said to have any basic unit of analysis at all […]

Methodological pluralism will arise from the plurality of objects under study and various types of evidence will receive acceptance if the object allows and if an advance of relevant knowledge can be extracted.

As a perspective and a functional one, pragmatics asks What is it to language use? by asking What does using language do for human beings and what do human beings do with language? It is thus part of action theory, where modes of action (one specific mode where others could have very well taken its place) are described and explained.

Wrong? No. Right? No. What else?

The first arguers out there were, after every definition, quite obtuse. I am thinking of the Milesian School. And it is not as if they were wrong in their approach, ignorantly wandering them fields unbeknownst. In fact, if we judge their capacity of getting along with what they had, they were quite sharp. So what makes them stupid?

Well, they were what we today might call physicists. They were not interested in the why’s and the what now’s but the what-of’s and what-from’s. That, at least, is the image we get from Aristotle’s depiction of their interests – the only one unfortunately. And it is as physicists that they appear to us as rather narrow minded. Otherwise, it is hard to see something wrong with the argument that the world was made in the same way tempests and lightning is born in the middle of the sea: the wind, Anaximenes tells us, is sometimes captured inside clouds; it then fights and struggles until it manages to crack one of them open, producing that violent, formidable clatter called thunder. Rather poetic but, as we now know, this is what happens.

And everything went like this. Men at the brink of a continent, pushed away by the Persian empire more and more into the sea, for them more or less everything must happen in the way natural phenomena happens at sea. The formation of the Earth, that of the stars, neither should be much different. Doesn’t water evaporate, producing clouds which produce rivers which brings alluvia, with all its life and support. Doesn’t it follow, then, that everything is made, literally, “out of” water?

We might appreciate this lack of “oh, Ol’ Daddy and Ol’ Mummy did every blessed thing around here!”. But if we do we should also be ready to evaluate these arguments. How in the world could one do that? What are the evaluation criteria for arguments which are blatantly false, in an epoch where they were strikingly obvious? And even assuming we find such criteria (with a little help from historians and great commentators of ancient text) what will the decision look like? If it is a confident “a-ha!” in the face of empirical error, then the analyst should settle for nothing more than being the uninteresting executioner of false beliefs. If it is a detailed and qualified “welly-well, here’s how you should view it”, then the analyst should be prepared to counter the accusation of relativism.

Now, the Milesian School aside, in The Poverty of Historicism, Karl Popper meticulously puts forward a bunch of arguments for why neither of the two is the right thing to do with respect to historical judgements of this sort. Well, maybe not of this sort precisely, but with ones having to do with social phenomena, judgements which, one should assume, are just as bad in the past as they are now. But if the right thing is somewhere in-between: Who decides where this right spot is?

When you write like this

Among the easily discernible differences between medieval and post-mediaeval philosophy there is a striking difference in forms of literary expression. For one thing, whereas the mediaeval wrote in Latin, in the post-mediaeval period we find an increasing use of the vernacular. It would not, indeed, be true to say that no use was made of Latin in the pre-Kantian modem period. Both Francis Bacon and Descartes wrote in Latin as well as in the vernacular. So too did Hobbes. And Spinoza composed his works in Latin. But Locke wrote in English, and in the eighteenth century we find a common use of the vernacular. Hume wrote in English, Voltaire and Rousseau in French, Kant in German. For another thing, whereas the mediaeval were much given to the practice of writing commentaries on certain standard works, the post-mediaeval philosophers, whether they wrote in Latin or in the vernacular, composed original treatises in which the commentary-form was abandoned. I do not mean to imply that the mediaeval wrote only commentaries; for this would be quite untrue. At the same time commentaries on the Sentences of Peter Lombard and on the works of Aristotle and others were characteristic features of mediaeval philosophical composition, whereas when we think of the writings of seventeenth-century philosophers we think of free treatises, not of commentaries.

(Copelston, 1994, p. 5)

People with a propensity for literary criticism & cognitive sciences, if they’re out there, should wonder how the predominant genre of an epoch affects the predominant beliefs. Hard to isolate, maybe, but one cannot help thinking that when you decide to write in commentaries you decide more than just the spacing and the subject of your sentences. Amongst other, maybe more subtle, things you also choose the amount of personal input you are expected to produce.

The equivalent nowadays? Is there an equivalent nowadays? Are the scholarly genres of the twenty-first century rigid and obstructive? Maybe not. At least not apparently. But think of the academic paper. Who decides what counts as and what cannot count as a scholarly paper? And on what grounds does one make this decision? Isn’t novelty amongst the criteria? Could you possibly get through with a paper that brings up nothing novel? So then, aren’t you forced – the word might be harsh, but you are, in a sense, forced by the context – to bring about something new if you want to get published? This might be so. And not only in the natural (hard) sciences, where even if you are engaged in testing an old hypothesis, you are doing so after a new fashion (with new data, by a new method etc.) It is also the case, maybe even more so, in social sciences and, ultimately, in philosophy. Post hoc ergo propter hoc? Possibly, but here’s a nice research idea for an argumentation theorist: How authors manage to convey the message that their paper is new, i.e. brings something novel, to the field in which it is produced. Strategic manoeuvring, straight-out argument structures, linguistic devices, implicitness – the whole deal.

Aye, aye, capitan! Ad Verecundiam it is

There is, I think, a muted historical side-note to every discussion on ad verecundiam. As neat historians will tell you, philosophy and science have risen precisely against blind faith in authority. To take a conveniant example of both in one,Thales didn’t say everything is made out of water simpliciter or that everything is made out of water because Zugma-Bugma said so. As was the Milesian School’s wont, they argued – poorly, but cunningly – for what they claimed. “Water is best!” was Thales’ one-liner, but it did not end there. Russell adds, not without a touch of irony: “The statement that everything is made of water is to be regarded as a scientific hypothesis, and by no means a foolish one. Twenty years ago, the received view was that everything is made of hydrogen, which is two thirds of water.” (Russell, 1946, p. 44). And that to say nothing about the geometrical, utterly non-religious proofs attributed to Mr. Thales. Ta-naa!

But in a way argumentation theory also had its incubation in the revolt against authority – although an authority of a less religious stripe. The problem of authority in arguments is thus pressing: if, in arguing, we sometimes appeal to experts, and if, in appealing to experts, we are not invariably committing a fallacy, where do we draw the line? Despite its immediate applications (“tell me where the line is and I’ll tell them if their appeal to authority is fallacious or not”), this is a very philosophical issue. I say “philosophical” because of this question underlying the abovementioned one: What makes an argument fallacious?
Really? Are we still floundering around this question? I think so. Pointless? Maybe. Instructive? Sure. Engageons-nous!

The problem of part-time fallacies is that you have to identify a class (The X’s) and a subset within this class (The “Bad” X’s) without saying to much about the specific socio-cultural circumstances pertaining to the evil subset. To do otherwise would be fishy: if we judge the fallaciousness of arguments by their “field”, then relativism and ultimate futility are not very far. I keep putting “argument” in italics because of the solution found by Walton (1995; 1997; 2006a). As his pragmatic theory of fallacy led him, Walton found a cute way out of the dilemma of appeal authority. First, he makes distinction in the meaning of “Bad” from “The ‘Bad’ X’s”: there are weak­-Bad and fallacious-Bad.

Many of these cases are not fallacious ad verecundiam arguments, but are weak arguments based on appeal to expert opinion. (2006, p. 754)

Unfortunately, the distinction is only gradual and he continues with showing how some weak arguments are “not so bad that [they] should be called fallacious” (754). But anyway, blurred as it may be, the category of “fallacious-Bad” appeals to authority is there. Thus, assuming there are some clear-cut cases which exhibit the essential properties of an appeal to authority which is fallacious-bad, what would such a case look like?

The fallacious cases seem to be the ones where the arguer who appeals to expert opinion tries to preclude or shut down the examination interval. The fallacious cases, according to Walton (1997), are ones where the arguer tries to block off or shut down the possibility of critical questioning of the appeal to expert opinion through the use of certain argumentation tactics. Often these tactics are of a pre-emptive sort, showing that the proponent of the appeal to expert opinion is not really open to critical questioning at all, even though she may give a surface appearance of being so.

This, however, is not a solution to the problem we were formulating. Remember the italics? We are interested in the fallaciousness of the argument from expert opinion, not in the fallaciousness of the moves with which the arguer tries to preclude or shut down the examination interval. If the analyst identifies some speech acts that might fit the description, a “blocking move” as it is termed in (Walton, 1997), then those speech acts, not the ones constituting the argument, are fallacious. If the authority is invoked as the last word on the subject, then it is the linguistic means of conveying this that are fallacious, not the argument itself.

Thus, if not by blocking a legitimate shift to an examination dialogue, when are appeals to authority fallacious?
There are certainly many things that can go wrong with the authority. The classic procedure is to identify a scheme, a couple of critical questions and a set of criteria for judging when those critical questions are observed. The student is advised to do this: take the raw text, identify the scheme, formulate the critical questions, see whether there are well dealt with, judge.

But notice, again, the more philosophical question underlying this advice: The only way an argument can be fallacious is by not having its critical questions well observed. Thus, a fallacious argument is a very very very bad argument. This seems also to be the position of the pragma-dialectical approach (Eeemeren & Grootendorst, 1992; Wagemans, 2011). Only a couple of the rules of the critical discussion apply to the argumentation stage, and those rules tell us about the argument scheme and whether the critical questions pertaining to it are well-answered. In his Appeal to expert opinion (1997), Walton himself is at pains to give the most accurate description of the scheme and a well organized taxonomy of all the critical questions that might relate to that scheme (with sub-questions and everything).

The expert might be irrelevant, he might be biased, wrongly quoted etc. This, however, does not guard us against “The fallacy of appealing to the opinions of the doctor who secretly wants your grandmother dead because she is maddly in love with your grandfather”.

Assuming, however, this sort of atomism is not around the corner, one cannot help asking … is this it? “A fallacious argument is a very-very bad one”? If not, how else can an argument be fallacious, i.e. against the rules?