GRICE'S DISIMPLICATURE

Speranza

I would like to reflect on Peirce's theory of signs.
erhaps Peirce will help us analyse what Sidonius meant by inplicatura [sic] and what I mean by disimplicature.

Peirce was, as he says himself, “a pioneer or, rather, a backwoodsman” in the study of the subject of meaning, or "semeiotic," as he pretentiously called it (he was from New England!)

It is not very surprising that what Peirce, otherwise a brilliant philosopher, has to say about implicature and disimplicature is complicated and obscure when one recognizes that his pioneering achievement was the work of a highly individual, subtle, acute, though -- let's grant it -- not very systematic or disciplined mind.

But neither is mine!

Neither is it surprising that Peirce's writing is punctualised by penetrating and suggestive observations about stuff, including the fact that it's best to speak of an utterer IMPLICATING that p, rather than 'implicature' per se, and mutatis mutandis, an utterer DISimplicating that q, rather than disimplicature per se!

If you get my drift.

Partly because many problems and doctrines of contemporary Oxonian interest can be traced back to, or at least found in, Peirce, and partly because Peirce's own writing is nowadays (at any rate at Oxford, today, under Austin's bad influence -- his playgroup!) unduly neglected, I propose to consider some of the things Peirce says about "sign" a starting point for my discussion of the subject of "meaning," "implicating," and, notably, "disimplicating."

I shall first consider, as briefly as can without injustice, Pierce's views about the theory of signs in general; and then go on to consider the "classification," or taxonomy as I prefer, that Peirce worked to make introduce within the class of “signs”

********************************************

Peirce's GENERAL "theory" (or 'analysis') of signs (what he called “Semiotic”) -- is this a feasible enterprise or is yet again a philosopher craving for generalities?

(For the record, Peirce was born in 1840 and died in 1914. Taught only a little -- although perhaps "a little too much"!).

Peirce’s General Theory of Signs

Peirce appllies the lexeme ‘sign’ to a great variety of subjects, including many items which would NOT, in ordinary usage, be called signs at all.

At least by Austin!

This departure or extension of the use of ‘sign’, it should be noticed, is not of course necessarily to be condemned. At least by me!

For it may be that Peirce's extension on the use of 'sign' draws attention to some important common feature which a great variety of items have share -- if not a Humeian 'essence'! Something more like a Witters-type 'family resemblance'.

As a random selection from the items which Peirce calls "signs" we may list:

-- the rolling gait of a sailor
-- a diagramme
-- a shout of ‘Hi!’
-- the word ‘that’
-- any significant sentence
-- a weathercock
-- a picture of something

Many of these would NOT ordinarily be called "signs" by Austin!

It is important, for an understanding of what Peirce is doing in discussing the lexeme "sign", to realise that there is good reason to think that Peirce would have regarded the PREDICATES

“... is a sign”

 and

“... has a meaning”

as, I hope, interchangeable!

At least "has a meaning" is good Anglo-Saxon!

But 'sign' has a long Roman history! (As Eco and his disciples well know!)

In support of this, I need refer only to two passages.

In the first of these passages, Peirce argues that the verb ‘to mean’ stands for an irreducibly triadic relation (as Sharpless has repeatedly pointed out to me!).

In a second passage, Peirce argues that the sign-relation is irreducibly triadic.

One’s overwhelming impression is that the same point is being made in both passages.

Which is not bad, since I like to repeat myself, too!

Now, in ordinary usage, such an expression as

“... has a meaning”

(or as I prefer, "I mean...")

is of wider applicability than

“... is a sign."

But not all of the items which Peirce calls a "sign" can be said to ‘mean.' E.g.: a picture, possibly also a diagramme.

So, if Peirce equals

"has a meaning" with being (in his "sense") a sign, this involves an extension of the use of ‘meaning’.

Perhaps Peirce was not an Anglo-Saxon purist as I am!

I propose now, as briefly as I can, to indicate in more detail the extent to which Peirce broadens the ordinary use of ‘sign’ -- i.e. how he departs from it -- anathema for Austin!

It will be convenient to consider separately two ranges of items.

The first rage are those of a linguistic (or quasi-linguistic) character: roughly speaking, those which are or may be used in making a communication. ("I mean that p")

The second range are those of a wholly non-linguistic character. (Although my avowed aim is that rationality is monosemous!)

This rough-and-ready division I hope (if time permits -- as in Oxford it allways does!) to replace by a more precise one.

Let us take the second range first! (Second things first!, as my Oxford tutor at Corpus would say!)

Within this class of items what would ordinarily be called a ‘sign’ (at any rate a sign ‘of’ or ‘that’ …) would seem to be the so-called ‘NATURAL signs’ (that Hobbes identifies, following Ockham -- "circulus in taberna significat naturaliter...").

And it is, at first sight, tempting to assume (roughly) that, if I may use a material conditional:

****************

whenever the presence or existence of "x"
is a ground for supposing the presence or
existence of "y," "x" is a "sign" of "y" for an interpretant "z"

M(x, y, z)

****************

In brief: x =means y, for z.

But this would be wrong for on a number of counts.

People may perhaps come to accept a particular house as being my house on the strength of its having a green door.

I.e. they will recognize my house by its green door -- it's English-green, actually, not Irish-green!

But it would be UN-NATURAL to speak of the green door as being a SIGN OF my house, or to speak of the fact that my cottage on Banbury Road having an English-green door is a "sign" that it is my cottage.

Cfr. Ockham's problem about the alley-way entrance and the food.

I think the principle (or rather one principle) involved here may be inadequately stated by saying, e.g., that (where ‘x’ and ‘y’ refer to individuals) we do NOT speak of "x" as being the sign of "y,"  unless we are prepared to allow that this can fall under some general CONSIDERATION to the effect that things of a kind or class to which "x" belongs are generally associated with things of a kind or class to which "y" belong.

But we are not here able to say that e.g. houses which have English-green doors (and which contain other unspecified characteristics) are always or usually mine. I wish!

The other type of case suggests that some such principle as the one I have indicated may not be the only one involved in the preceding scenario.

For though not only do I NOT speak of the presence of a feature by which I distinguish a particular individual bird from other birds as a 'sign' (OF the presence of that bird), I do not speak, either, of the feature by the aid of which I distinguish or recognise a particular kind of bird as a 'sign' of the presence of a bird of that particular kind.

E.g. I do not or would not speak of a bird’s having a pink beak as being a "sign" THAT it is a "Lesser Nitwit," if you have ever seen one.

And here of course I can produce the general consideration which applies to this case:

Birds with pink beaks (plus certain unspecified further characteristics) are always (or usually or often) Lesser Nitwits.

Now, it seems to me that we tend to distinguish in our talk the features by which we recognise something as being of a certain kind (by which we distinguish so-and-so’s) from a ‘sign.'

A 'mark’ would be a more appropriate label than a ‘sign’ in such cases.

Marks (most naturally) ‘distinguish.'

Signs (naturally) ‘indicate.'

But such a distinction does not always seem to be made systematically.

And, in any case, one would desire some further account of the _basis_ for this distinction between a 'sign' and a 'mark.'

This I fear I cannot supply at the moment, or perhaps never!

Maybe some of the following cases may show the principle governing some of the following types of case still to be mentioned and it all might prove relevant.

"... means" but not ‘is evidence’ seems appropriate here.

I think we do not in general speak of x as being a "sign" of "y" when the existence or occurrence of x is a constitutive condition of the occurrence or existence of y.

E.g.

I doubt if we should say that a man’s having lost his temper is, on a particular occasion, a "sign" that the man is irascible (though it might be _evidence_ that he is irascible).

For being irascible consists in having a tendency to lose one’s temper.

A slightly somewhat different case involves a value term like 'good.'

We should NOT want to say that the fact that, in general, a 'well-reasoned' essay is a "sign" that the essay is a "good" essay for analogous reasons.

Now replace 'sign' by ‘mark’:

“It is, by definition, one of the marks of a good essay to be well-reasoned."

I think we tend NOT to use the word ‘sign,' at any rate without qualification, of x, if the presence of x leaves no room for doubt as to the presence of y, but rather, we tend to use ‘means’ or -- if one feels like following Witters -- ‘shows.'

We would NOT normally say of a particular envelope that its bearing a post-mark is a "sign" of its having been through the post, because in normal circumstances this would be taken as establishing that it had been through the post.

However, if special circumstances DO create suspicion that e.g. the post-mark may not be genuine, then perhaps we might be more ready to use ‘sign’.

And I've seen such cases, my friend S. N. Hampshire used to live in W2 but when posting his letters he would go to W1 -- yes, he IS pretentious!

In this connection, we may note that the form

“x is a "sign" that p, but possibly ~p”

is legitimate, while the forms

“x means that/shows that p, but possibly not-p”

are not.

In other words, never forget the test of cancellability or what the Kiparskys call 'factive'!

On top, we do NOT normally speak of x as being a sign of y if the existence of x is a cause, the explanation, or part of the explanation, of the existence of y.

E.g.:

We do NOT say

The death of Mr Smith is a "sign" that Mrs Smith will be unhappy."

And, oddly, we do not say this even if we are uncertain whether Mrs Smith will care.

We do NOT say

“His catching bulimic plague is a sign that he will die soon."

But of course there is no ban of speaking of the effect as the sign of the cause.

I think we withhold “... is evidence that p” here too.

We can however say ‘... means...’ or ‘... probably means...’.

Additionally, we do would NOT normally say e.g. that the fact that the first ten people to arrive at a meeting were wearing gowns is a SIGN (but a prediction) that that the next person to arrive will be wearing a gown, even though we may think that this fact makes it probable that he WILL be wearing a gown.

That is, we do NOT speak of a "sign" in cases which are specimens of what logicians sometimes call analogical arguments or inferences -- and Peirce, to be different, would call abductions.

Nor, incidentally, do we speak of the above mentioned fact as being a "sign" that everyone at the meeting will be gowned.

Since the collection of these somewhat miscellaneous-looking examples of things that we would NOT call a "sign", it has occurred to me that it might be worth considering whether there may not be a quite simple positive requirement for a "sign" which would EXCLUDE all those scenarios from the status of a "sign." Namely:

For x to be a sign of y it is necessary, though maybe not sufficient, that y should explain x,, wholly or partially.

This requirement, if authentic, eliminates the cited items from being a "sign"; and would also explain further features of our use of the lexeme "sign," e.g., why we might speak of the fact that a man has a stoop as being a "sign" that he is old, but not as being a sign that he has grey hair or false teeth. For we would allow

"The man has a stoop because he is old." but not

"He has a stoop because his hair is grey or his teeth are false."

The obvious apparent difficulty over this suggestion is that we often speak of refer to, say, a present event as a "sign" of a future one.

And it may be asked, how can a future event explain a present one?

In answer to this, one must say that we certainly do say such things as

"The rope creaks because it is about the break."

Or:

“His voice grows faint because he is about to die.”

Such utterances may or may not raise a problem for the philosopher -- at least not to me -- never mind Peirce -- perhaps he wasn't a philosopher at heart!

But they are undoubtedly legitimate ways of talking here at St. John's!

Other requirements to be fulfilled by a "sign" have been suggested.

E. g. that they must be more obvious or accessible than their "significata,"

Or:

that they must be in spatial or temporal proximity to their significata.

I am far from convinced that such requirements are authentic.

That they might seem so might be explained what plausibility they may have might be explained by pointing that they are usually, if not always, fulfilled in the case of items which fulfil my own suggested requirement for a "sign."

In any case, I do not purpose to discuss this further.

What about the linguistic or quasi-linguistic items?

In this field, the following seem to be the most obvious types of a "sign" case in which the word ‘sign’ is applicable.

(i)

Certain non-verbal (or not wholly verbal) but conventional communication devices, primarily at least to be visually apprehended: e.g.

‘+’ (addition sign) "^".

Both these mathematical signs might also be called "symbols," though it does not seem that perhaps not all such symbols would in fact also be called "signs."

-- road signs, e.g. level-crossing sign.
-- shop-signs, e.g. barber’s pole.
-- a conventional sign such as a map of the Oxford colleges.

Gestures, whether conventional or not, used as communication devices (obviously to be visually apprehended).

Nota bene in neither of the above types of case do we naturally speak of the items as ‘signs of’ or ‘signs that’ though we may speak of them as ‘signifying’ or’ ‘signifying that,' as Cicero would delight in!

Actions, normally at least usually, though not always, non-conventional, used as communication devices e.g.

“The chief took his hand as a sign that all was forgiven."

“They shot down the German aeroplane, as a sign that they would stand no nonsense.”

Nota bene that the natural phrase is ‘as a sign’ may be followed by ‘that’ or ‘to.'

These may often be called ‘gestures’, in a metaphorical sense.

No doubt there are other categories of Class a signs.

But these perhaps are enough for the moment, I hope.

Note particularly a word (for all Locke said -- vide my "Studies in the way of words" versus his 'way of ideas' and 'way of things'), a phrase, a sentence (even a statement) is NOT a sign. (Although we should perhaps re-read what Cicero said of Aristotle's semeion!

The utterance, use, or making all of these may be a sign, though! (And this was Locke's point -- cfr. his example of Prince Maurice's parrot).

 (Class b) (e.g. of gullibility or intelligence) but is not a sign (either class a or class b) of what the word (phrase, sentence) means nor of what the statement itself asserts.

It's now high time to return to Peirce and his krypo-technical jargon!

***********************************

Peirce’s Theory of Signs

Peirce’s general theory of signs is both fascinating and obscure.

I hope that the following attempt to simplify a welter of pronouncements will not prove too much of a travesty, but a rhapsody on his theme.

What seem to me to be the salient points of Peirce's "theory" (if not analysis) may be summarised as follows.

First, to be a sign (to mean something) is to stand in a triadic (three-term) relation:

M (x, y, z)

The initial move in attempting to show this is to claim that a sign s for s to be a sign is for s to be a sign of something/some object (o) to somebody -- the interpreter, or addressee.

The relation in question is not merely triadic but ‘genuinely’ triadic, i.e. not reducible to any “complexus of dyadic relations." (Peirce seems like he had been reading too much of Russell!

Contrary to what one might expect, the terms of the sign relation are not Sign (s), object signified (o) and a third or organism, but sign, object and what Peirce calls ‘interpretant’.

The following expansion suggested by W. B. Gallie seems to fit at least some of what Peirce says about signs.

S is a sign provided that, i.e. only if, s evokes (in some mind or organism) a response (‘interpretant’) appropriate to o (object).

Aain contrary, perhaps, to what one might expect, the meaning of a sign S is not the object o for which it stands, but rather its interpretant, or at least rather  -- vide later writings -- a certain sort of interpretant).

For s to be a sign, its interpretant must be mental.

And so (according to Peirce in his earlier writings at least) a thought.

And so (again) a further sign -- since all thought on Peirce’s view consists of signs!

Further it must be a sign of the same object as is the sign of which it is the interpretant.

Gallie takes this as being ensued to be the case given that the interpretant response is to be appropriate to o.

If appropriate to o, itself a sign of o.

This entails that every sign generates an infinite series of interpretants all of which have the same Object.

Peirce is perfectly willing to accept this consequence.

A further qualification mitigate meets the obvious objection that it would be logically impossible for there to be an infinite series of actual responses (interpretant).

Peirce allows that the interpretant need not be an actual response, but may be merely a potential one, i.e. one which would occur given certain further, in fact unfulfilled, conditions. He could be Platonic!

Some further elaborations of Peirce's position may be mentioned.

But this abrupt summary will do for the moment.

Some comments are now called for.

We should ask what Peirce’s theory of signs is a theory of, in view of the obvious fact that ‘sign’ is not here used in its everyday Oxonian "sense."

One reply might be that it should be regarded as an attempt to elucidate the ordinary concept of meaning.

On this view we should have to assume (as indeed is no doubt the case) that Peirce has taken the concept to have a somewhat wider application than it in fact has, e.g. has omitted to notice that pictures or maps would not ordinarily be said (as such) to ‘mean’.

Now it will undoubtedly be useful, if one is interested in the analysis of ‘meaning’, to consider Peirce’s theory in this light, and I don’t think he could complain.

Ideed, I shall often take this line myself.

Nevertheless only to do this would be perhaps unfair to Peirce.

We should I think also take note of the fact that Peirce, in talking about ‘signs’ is talking about a very diverse collection of items, which nevertheless we (as will he) feel it fairly natural to group together in one list (even though there may be no obvious ready-made title to put at the head of the list); and what he says might tell us why we would readily list them together, what they have in common.

The two approaches are not incompatible.

For such items with meaning (in the ordinary sense) form a sub-class of Peirce’s signs, any condition required to be a sign in Peirce’s sense would constitute a necessary, though not sufficient, condition of having a meaning.

One caveat:

We must not assume, as Peirce’s account would suggest, that there is some one common general character which all the members of the list share.

The members of the list might merely be linked by what Witters once called (indeed, more than once, if we trust Anscombe!)  “family resemblance”.

Indeed, if one looks for any single character shared by e.g. pictures, maps, sentences and weather-cocks, it is not easy to pick on one.

The best I can suggest is that they are all items from which something might be learned about some further item.

But whether this delimits the class of Peirce’s ‘signs’ I do not know -- perhaps Witters or Anscombe do!

Second, I do not propose to spend long here over Peirce’s thesis that the ‘sign-relation’ (meaning) is a three-termed (and irreducibly three-termed relation).

If all that were meant were that e.g. signs do not have meaning "in vacuo," that if there were no people to respond to signs, there would not be signs with meanings, perhaps no one could object.

But in fact Peirce clearly means more than this.

He seems to have meant inter alia that to a statement of the form

‘x meant/means y’

the response

'to whom?’

would ALWAYS be appropriate and have an answer.

But if this is his view,

it is false.

Counter examples include:

“The evolving of the earth meant the emergence of life."

“Kant’s words meant …” (choosing some passage which you think has hitherto baffled every one.

It may be misleading in so far as it suggests that the meaning in general, and so (naturally) the meaning of communication devices, is to be thought of in terms of the addressee to which they are or might be addressed (so encouraging a causal theory of meaning in the sense of meaning appropriate to communication devices) rather than in terms of the utterers or users of the devices (which would encourage an ‘intention’ theory of the type I favour).

As I prefer an intention theory to a causal theory, I regard this as unfortunate.

The main difficulties in Peirce’s begin with the lexemes ‘object’ and ‘interpretant’.

A word about the Object of a sign.

As already mentioned the Object of a sign is not what the sign means.

We mentioned the suggestion that Peirce thinks of the Object as that to which the response is to be appropriate.

Peirce does not actually say this.

Indeed he does not say a great deal about the Object.

He does say however, that the Object is “that which the sign presupposes an acquaintance in order to convey some further information about it”, that is what an addressee has to be in a position to identify in order to comprehend the information conveyed by the sign (which is information about the Object).

Cf. Peirce’s example of the man, looking at empty sea, who is told “that vessel there carries no freight at all”.

Roughly to comprehend the information so conveyed he has to know what vessel is being spoken about, even if, as may be, there is no actual vessel present there.

Roughly then a the Object is that which the sign tells about.

This is not inconsistent with the earlier suggestion.

The main trouble is with ‘interpretant’.

The interpretant seems to be regarded by Peirce as being both a response (wider sense) the making of which (or the disposition to make which) is a manifestation of understanding a sign (taking it as a sign of something) and as a further sign which elucidates or develops the original sign.

This suggests strongly to me that Peirce is operating with, and conflating, two seemingly different views on meaning.

One is that a sign having meaning consists in its

On both views a sign having meaning would consist in its being capable of being understood.

On the first view this would in turn ‘being understood’ would consist in the sign’s capacity to produce producing in an hearer audience some effect or response (or a disposition to which may be ‘potential’—i.e. in a disposition to do certain things), this response being of kind appropriate to the Object of a sign; on the second view the signs being capable a sign being understood would consist in the audience giving, or being able to give, some translation of or elucidation of the sign.

A conflation of these yields the view that to understand a sign is to respond to the sign by giving (or being capable of giving acquiring a disposition to give) a translation or elucidation of the sign.

Now such account of meaning taken alone has its difficulties: the first (roughly a ‘causal’ theory) will, I suspect, turn not to be unsatisfactory (however it may be hedged around) when the meaning of communication devices is considered.

I do not think the sense of ‘mean’ in which utterances can be said to mean can be satisfactorily defined in terms of the [responses/reactions] they arouse or tend to arouse.

The second I will suspect (that having a meaning = roughly being translatable by a further sign) is open to three obvious objections as a definition of ‘meaning’

(i) that an indefinite series of ‘translations’ for (linguistic) sign might be provided providable without its being the case that the sign means anything (e.g. if I invent a ‘nonsense’ language -- call it "Deutero-Esperanto".)

Generally speaking in order to be asserted that a (linguistic) sign means something (or to know what it means), one has not merely to be able to relate the sign to other signs, but at some stage to relate the use of the sign to non-linguistic situations (to the world).

(ii) that as a definition of meaning it would be circular (and one would think objectionably circular), since it explains the meaning of signs in terms of (roughly) other signs having the same meaning the existence of other signs having the same meaning.

(iii) And the conflation of the two views (apart from any objection to such of the conflated items) seems objectionable in that in so far as it concerns the implication that for a response to be a manifestation of understanding it must consist in giving (or being able to give) some translation or quasi translation of the original sign.

But obviously I can show that I understand a sign by the way I behave (as Peirce himself later admits in the case of the soldiers response to an order) where the behaviour does not consist in providing any sort of elucidation of the sign which I understand.

If Peirce has made this conflation, one or more of the following factors may have assisted him towards doing so.

(1) The ambiguity of the word ‘interpret’ which may mean (a) ‘make sense of’, ‘understand’ (b) ‘explain the sense of’.

(2) A possible haziness about the meaning of saying (as he does) that the interpretant (response) must be ‘mental’ (must be a thought and so a further sign).

This might be understood as saying either that the interpretant response must be a response which counts as ‘thinking’ (in the sense roughly of ‘cogitating’) or that the interpretant response must be characteristic of an organism with a mind (i.e. roughly must be a manifestation of intelligence, or of a certain order of intelligence).

On the second interpretation the statement that the Interpretant must be mental is at least plausible; but of course on this view it need not be a piece of cogitation.

On the first interpretation if the original statement were true (and cogitation consists in using signs) then the interpretant would have to be a sign (linguistic or quasi linguistic), but on this interpretation the original statement doesn’t seem to be true.

A failure to distinguish the true possible interpretations might result in the acceptance of the statement (“interpretant must be mental”) understood in the first sense because of its plausibility when understood in the second sense.

A possible unconscious shift between a wider sense of sign (to include ‘natural’ sign) and a narrower sense of sign (to include only invisible or quasi linguistic devices).

If a response to sign is appropriate to certain Object, then maybe it is plausible to suggest that it would be a sign (natural) of that object.

There might then be an unjustifiable transition to supposing that its response it would be a sign in the manner sense of that object, and so an elucidation of the original sign.

I suggest Peirce might (as regard general discussion of meaning) be not too unfairly represented as having moved roughly as follows

He seems for start to combine two different views, without saying he is doing so.

On both s has a meaning = s is capable of being understood.

But then 2 possible views about analysis of being understood. s understood by x? = (1) some effect on x produced or capable (given certain conditions) of being produced on x (effect being in some way appropriate to o (object of sign) (2) some explanation or translation given or capable of being given by x of s.

This explanation (definiens) what (in case of language any way) turns out to be called logical interpretant So s has meaning seems to be (in Peirce) (1) s is capable of producing effect on interpreter (2) s [is] capable of being explained If I’m right how could then [‘]has meaning[’] be explained (1) Natural to suppose effect “thought”, + coupled with view that all thought [is in] sign and “mostly of nature of language”11 might well lead to 2 (thought = interpretant of sign)

(2) ambiguity of interpretation (= interpreting, result of interpretation) cf. analysis definition

(3) that any sign [is] explainable [is] something that seems plausible and natural to want to hold it true (hope at least it will be consequence of analysis of ‘meaning’).

But this different from being analyzed Peirce then gets its knives about regress of interpretants: series of interpretants which are signs must be ended by final “ultimate logical interpretant” which is not a sign (or at least “not a sign in the same way as that of which it is interpretant is sign”).

This turns out to be a “habit” (and also a “significate effect of sign”)

Lastly the trouble/difficulty about being imprisoned in language (but this would seem to require rule of special or different kind; in fact Peirce provides them with operation on sensible result (2) circularity difficulty about 2nd analysis of meaning (not mentioned but perhaps Peircevaguely disturbed about it).

We now seem to have Peirceagain running both lines (provided special kind of rule to tie language to experience[)] + at same time saying that won’t quite do since “it doesn’t tell us of essential nature of effect on interpreter brought about by semiosis of the sign" [Semiosis explained as action or influence involving the well known triad] Qualisigns, Sinsigns, Legisigns.

You will remember that a sign was defined as something that stands to somebody for something in some respect or capacity.

Peirce with his eye on this definition now proceeds to distinguish three different trichotomies of signs.

The first trichotomy depends on the status of the sign itself,“whether it is a mere quality, an actual existent or a general law”.17 If it is a quality he calls it a “quali-sign”, if an actual existent he calls it a “sinsign” (or “Token”), if a law he calls it a “legi-sign” (or else “Type”).

I shall first consider the last two of these three divisions. You may well be familiar with the type-token distinction and I think Peirce himself explains fairly clearly how to apply the words “type” and “token”. [Read of 280–281]

The following comments seem called for on what Peirce says: (1) In his description of a type-sign or legi-sign as a law he seems to be both confused and to be struggling with an inadequate terminology.

He is confused because often having said that a legi-sign is a law he goes on to say that the law is usually established by men and that every conventional sign is a legi-sign but not conversely.

He clearly seems to be confusing here a law equals rule for the use of a sign with the sign itself.

This confusion may be due to inadequacy of terminology that I mentioned since I think suspect that what he is trying to say is roughly that a type-sign is an abstraction or logical construction out of its tokens and presumably to say something about a type sign, in his view, would be to say something general about the tokens of it, hence perhaps “law”.

There is obviously something in this but I am rather doubtful if he has it quite right.

The sentence “It is impossible that this word (type the) can be visibly on a paper and be heard in any voice”, suggests that the token work the could, in his curious phrase, be visibly on a page. I suspect he means by this that there this [sic, is] a [p. 12] sense of the word “word” in which a word be just a “mound of ink” or a part of the page; though the meaning of the word “word” fairly clearly has something to do with such things as mounds of ink I am doubtful if this straightforward answer will quite do.

(2) It is important to notice that the type-token distinction applies to signs other than words.

Other examples would be in the case of maps and diagrams. Cf. the following sentence

"Have you bought the latest Ordinance Survey map of Oxon?” (to which the reply “No they have printed some more since I brought mine”, would be inappropriate) with “There are 20 o.s. maps in the building”. Now this distinction is of course useful in itself, but has it any bearing on a search for different uses of the word “mean”?

It certainly, I think, at least suggests an obvious though extremely important and much overlooked distinction.

The meaning of a word or expression is what is meant by a particular use of that word or expression.

This might be shortly described as the difference between a tenseless use of the word mean (e.g. father means male parent) and  a tensed use (e.g. “He meant by “father” the r.c. priest”).

I am fairly sure Peirce at least to some extent had this distinction in view. n.b. his phrase when speaking of a token “Such an event being significant only as occurring when and where it does”.

This suggests that a single word on a piece of paper in a drawer might not count as a token sign.

Though this distinction is important I think it would be wrong to assume that it is a type word that means (tenseless) something and a token word that means (tensed) something.

We should ordinarily speak, I think, of a type and meaning (tensed) something in a certain context, we should say We must notice however that the tensed use of “mean” is not confined to token words: we can say either

“The word “bank” meant something different each time”

or

“Each “bank” meant something different”.

The category of quali-signs is a highly suspect one.

I do not propose to spend much time on it.

A quali-sign is defined as “a quality which is a sign”.

Peirce says “it cannot actually act as a sign until it is embodied; but the embodiment has nothing to do with its character as a sign”.

“It is the mere quality of an appearance and is not exactly the same throughout a second”.

It seems to be identical with what he elsewhere calls a Tone where his example is “An indefinite significant character such as the tone of voice” where the sort of thing he seems to have in mind is a specific voice quality by which I recognise a person.

But obviously here I don’t know what it would mean in regard to this example that the “embodiment” has nothing to do with its character as a sign.

Another alleged example of a quali-sign is “a feeling of red”—presumably this means a sensation of red regarded by Peirce as a sign of the presence of a red thing.

It seems very difficult to think of examples of quali-signs which couldn’t be reduced to legi-signs or sin-signs. A possible candidate might be white as a symbol of purity (far-fetched). Terms (Rhemes) and Propositions (Dicent Signs) Two other division trichotomies said to be (a) dep according to relation of sign to its object i.e. depending on whether (roughly) sign has character as sign depends on its being like its object (Icon) [picture] or existentially related to that object (index) (e.g. smoke/fire) or its being specially related to the object the interpretant (symbols, e.g. conventional signs) [Details of this to be discussed about] (b) according as interpretant represents it as sign of possibility (Rheme, e.g. general word), as “sign of fact (Dicent sign = Dicisign e.g. indicative sentence) or “sign of reason” (Argument self-explanatory).

Examples I have given are from language, but divis trichotomy designed to apply to sign other than language. 2 trichotomies

(a) Icons or roughly “picturing” sign Indeces or roughly “indicating or pointing” signs Symbols or roughly “conventional signs”

(b) Rhemes or roughly “word-like” signs Dicents (Dicisgns) or roughly “telling” signs (telling that or to) Arguments—no explanation needed (Don’t se No ambuguity cases for [argument] to be symbol)

I will set out these in greater detail—beginning with last classification. Remember main object to consider different uses of “mean”, but of course we must inquire any light Peirce hung through on words allied to “mean” or indeed on any philosophical point reasonable germane to our present subject.

Rhemes (‘Word-Like’ Signs), Dicisigns (‘ “Welling” Signs’), Arguments First Rhemes “A Rheme is a signwhich,for its interpretant.

Does this mean anything different from ‘interpretation’?], is a sign of qualitative possibility, that is, is understood as representing such and such a kind of possible Object.

Any Rheme perhaps will afford some information; but it is not interpreted as doing so.”23 Other things said by Peirceabout Rhemes, but I don’t think will assist you if quoted. Examples of Rhemes—the (type or token) word ‘camel’ (and any general term)—the word “this (type or token), any Icon (e.g. a picture or map or diagram),—a spontaneous cry (a very queer one), “Hullo!”

Two obvious deficiencies (1) use of “qualitative” seems to cut out e.g. relation words (“about”, “beyond”) and adverbs (“slowly”, “speedily”) from being Rhemes. It is clear I think that he would want “qualitative possibility” not to exclude any descriptive word.

More important, what he says about Rhemes excludes one of his examples (“this”) from being a Rheme. This is neither a ‘quality’ word nor a ‘descriptive’ word, as Peirce himself is elsewhere well aware.

We must then not regard him as insisting on the “qualitative” however interpreted (e.g. to cover all descriptive words).

What can we take him to be saying a Rheme is?

Just any signs (i.e. roughly anything that means something, has a meaning) provided it is not Dicent (i.e. ‘telling’) or an argument; that is, tho’ it means something, it claims or tells nothing, nor expresses anything? (This I think part of what he’s saying but only part (ignores “possibility” altogether).

Even this of some interest for us, for (remembering our purpose of studying of ‘mean’) it suggests (what may be obvious but overlooked) that the sense or use of ‘mean’ as applied to sentences (‘complete utterances’) is different from that in which ‘mean’ applies to words (‘incomplete utterances’).

What more could he be saying? Let us remember that a ‘possibility’ is a ‘possibility that’.

Consider his own example of the word “camel”.

In what sense could the word ‘camel’ be a sign of that possibility that there are camels (that something is a camel)? Not in sense of asserting or informing e.g. me that (a) there are camels or (b) that it is possible that there are camels. No assertion (by definition) can be made by Rheme. I think he is trying to say what perhaps was not much more satisfactorily said by the people (e.g. I think Russell) that when we discuss expressions referring to universals, we should take as examples not ‘red’ (‘redness’) but ‘x is red’ (sentential functions).

Two points might be made in interpretation of this.

-- that we must remember that words like ‘red’, ‘camel’ (as any word) is subject to type or category rules, and to explain these is to explain how words can fit (significantly) in sentential (linguistic) contexts. You couldn’t know the meaning of a word without knowing how to obey type rules.

-- how important that any explanation of the meaning of word consists in explaining meaning of (senses of) sentences containing it. That implies I think that not only is sense of ‘mean’ applicable to words different from that applicable to sentences, but posteriori to it (to be explained is terms of it).

If so, this is important. (cf. those who find it a mystery, like C. S. Stevenson, in one place, how it is that words with separate meaning coalesce to form complex with meaning of its own (sentence): as if meaning of sentence were a sort of curious quasi-chemical compound of the meanings of the words in it) cf. Stevenson e+l p. 67: “… one of the most difficult problems that meaning-theory includes—that of explaining how separate words, each one with its own meaning, can combine to yield sentence-meanings. It is feasible perhaps to take each word as having a disposition to affect cognition, just as the full sentence does.

The problem reduces, then, to one of explaining the interplay of the dispositions of the several words, when taken conjointly.” [He then refers to analogy of giving magnets a relation between individual and group tendencies]

This suggest then that a Rheme is (or ought to be on interpretation of def) a sign which something which means something (a sign) and tho’ not itself a ‘telling’ or ‘claiming’ sign (Dicent), nevertheless can function as constituent in ‘telling’ or‘claiming’sign a dicent, + the sense of ‘mean’ applicable to it (Rheme) explainable in terms of that applicable to Dicent signs. But if this so then spontaneous cry (as Peirce says ‘sign (indexical) because it directs attention of utterer) a very odd example of Rheme.

It is certainly supposedly not Dicent in his sense (in so far as direction of attention This certainly doesn’t seem to derive any applicability of word ‘mean’ to it (I doubt if there is any applicabl from possibility of occurring in a ‘telling’ ‘complex’ sign. The most Rheme could mean here is “non-dicent sign” (i.e. only negative part applies). But a queer example altogether: does fact that if suppose I sit on drawing pins + shout, thereupon people look at me; does mere fact that shout makes them look at me render shout a sign (of me!) we feel inclined to say ‘no’.


Dicent signs There may of course well be non-linguistic Rhemes.

Peirce suggests icons in general; this I am not sure of; but what about conventional signs as maps, notes (♩) etc. in musical notation, e.g. naval flags (signalling), black dots (towns) in maps.

No doubt an elastic sense of ‘Rheme’ might have to be allowed; but at least many of these things rather like words in some respects. Dicent signs. In respect to language however, sometimes there seems to be yet a different sense of Rheme:

"A Dicisign necessarily involves, as a part of it, a rheme, to describe the fact which it is interpreted as indicating”.25 Here a Rheme seems to be not a word but as it we a sentence frame (?) (with gap-signs for any expression e.g. demonstratives—used for referring to refer to individuals). For there is no suggestion that dicisign contains more than one rheme.

This perhaps important only for interpreting Peirce.

Rheme then = (1) any sign which though not a dicent has ‘meaning’ in sense only explainable in terms of ‘mean’ as applied to dicents possible appearance as constituent of dicent signs (and therefore of ‘mean’ as applied to dicents) e.g. any word (2) sentence frame (?) with gap-signs for ‘referring’ expressions (3) any sign which is not a dicent (or argument) Dicent signs The right and general way of distinguishing dicent signs from other signs is by the fact that dicent signs are those which are informative or give information. But this is clearly only an approximation to Peirce’s full doctrine: for by this account only indicative sentences (among sentences) would be dicent.

Let me first quote some passage which supplements the first rough account of the meaning of ‘dicent’:

(a) “Symbols are by nature in the ‘indicative’, or as it should be called the declarative mood.

Of course they can go to the expression of any other mood, since we may declare assertion to be doubtful, or mere interrogations, or imperatively requisite.”26

(b)“By a Pheme [much the same as, a maturer version of, a Dicisign]I mean a sign which is equivalent to a grammatical sentence, whether it beInterrogative, Imperative, or Assertory.”27 [nb This does no doubt as Peirce indicates represent a development of his earlier notion of Dicisign]

(c) “A dicent is not an assertion, but [is] a sign capable of being asserted. But an assertion is a dicent. According to my present view … the act of assertion is not a pure act of signification. It is an exhibition of the fact that we subject ourself to the penalities visited on a liar if the proposition asserted is not true.

Cf. Buchler sel. pp. 103–10429 (d) “The Replica [Token] of the Dicent symbol is a Dicent Sinsign [Token] of a peculiar kind. This is easily seen to be true when the information the Dicent Symbol conveys is of actual fact. When that information is or gets real law, it is not true in the same fullness.”30

He then unfortunately goes on to say that a Dicent Sinsign cannot convey information of law.

As I am not concerned with Peircean scholarship, but with what can be got from him, I shall ignore this, which is due I think to a persistent assumption that gen general information can only be conveyed by type signs (not tokens); this is connected with his description of a type-sign as a law which I mentioned above].

The important point is the emphasis on the difference between Type + Token Dicisigns

Examples of dicent signals (?) signs: indicative sentences weather cock (+ natural signs usually) + photographs? combination of icon + index (e.g. painting) [I] Don’t proper have to talk about application of ‘dicent’ outside language: [I] reserve this for time when Peirce’s other distinction among signs more familiar obvious. Something funny about putting weather cock + picture in same basket. (8) Dicent (a) + (b) Peirce I think (a) saying that dicent (or Phemes at least—amended dicent) need not be indicative; imperatives etc. count as well (b) saying something more; something expressed by some philosophers saying that same proposition may [be] object of all sorts of different attitudes.

The A clear way of saying something like this found by R.M. Hare (Imper. Sentences Mind 4931)

Here his purpose no doubt different from Peirce + the philosopher in question.

The other two points which I think Peirce is getting at (with varying degrees of clearness) are both connected with the distinction between (a) what a sentence means (in general; timeless ‘means’; if you like, the meaning of (type) sentence) and (b) e.g. what I commit myself to by the use of a sentence (if you like, connect this with, or state it in terms of, token sentences). In discussion of these points ‘sentence’ will always be taken to be “non-misleadingly indicative sentence”; that is no question the sentences I shall be talking about will be indicative sentences and no such considerations as whether they really have e.g. imperative force are relevant. (though points I think apply just as much to the realm of imperative (optative) sentences as indicative).

In what follows here I’m under great general debt to P. F. Strawson who first pointed out to me the importance of the kind of distinctions here to be made, and applied it himself
in interesting cases (e.g. descriptions) from which I’ve learnt a great deal) I can’t suppose assume he would like my methods.

The first point concerns what Peirce is quite clearly I think getting at in quotation (c) [remind] above.

He obviously has in mind a pair of cases such as following: (a) one in whichI say sentence to myself (contemplatively) (b) where I say s it aloud (‘assertively’) in company (?) There is or may be a difference between what the sentence I use means and what I commit myself to in both cases the meaning of the sentence s is the same, but in the first case I commit myself (perhaps) to nothing at all, whereas in second case I do. But difference between committal and non-committal of oneself not a difference of meaning (ambiguity (?)) in sentences. I think his point can quite well be made in regard to such a pair of cases; but for my simplicity (+ extension of Peirce) I shall suppose we are concerned with two different ‘public’ utterances of sentences.

Suppose I find sentence (s)

“The whole of the French Riviera was in a state of panic on1st Jan1938” (a) in a history book (b) in a novel.IfI put a x in the margin of history book I shall no doubt be a nuisance but if I do so in the margin of the novel I shall be an fool idiot as well. It is inappropriate to criticize statements in novels (of the kind in question) on grounds of truth or falsity.

Nevertheless the sentence has one and only one meaning (is in no way ambiguous), nor has any private (special, real) meaning been attached to it by either writer. But the historian has committed himself, the novelist has not.

This example does however reveal to us a distinction in uses of ‘mean’; the sentence (s) means just one city; the historian wrote/said s and (presumably) meant what he said; the novelist said/wrote s but didn’t mean it. [cf. spoken cases “An Englishman, a Scotsman + an Irishman ....]

Neglect of the fact difference between “saying something which means _” and “saying something and meaning it”, neglect of the notion of assertion or self-committal may lead to queer views about sentences of fiction; seeing that T[rue] or F[alse] can’t appropriately be applied (tho’ they can be some sentences in ordinary life) may make we want to say that we have here different levels of language, a special sort of ambiguity, or might want to assimilate to ‘systematic’ ambiguity. [I have heard something like this been suggested]

Or if we are very very old fashioned, most ‘likely’ to say f, and not being happy about saying t, we might lean to fictional worlds of which fictional sentences are true. But novelists talk since quite unambiguously; only they don’t assert.

The second point or group of points are I think what p is rather dimly groping after in quotation (d) they concern The distinction between “what a sentence means” and what I mean by using the sentence (when [This already briefly alluded to] This very much a rhapsody on a theme by Peirce. But I think very very important.

The points concern distinction between “what a sentence means” and “what user of sentence commits himself (“what information sentence used to convey”—not the same tho’ may well coincide); only cases limit are those in which sp user doesn’t commit himself to anything, but commits hi does commit himself and what he commits himself to can’t be identified with what sentence means. [Often or always this may be represented as distinction between (‘what s means’ and ‘what s meant’ or ‘what x meant by s’) Ex. (a)

One obvious existence of difference is where use of s has special (different or fuller) meaning, this difference depending on the speaker’s (writer’s) use of some descriptive word in s. Too obvious to need clarification; but nevertheless great perplexities may arise from this kind of distinction.

E.g. R. M. Hare’s example of use of "good" by fossilized Indian Army officers.

This of course localized habit of use, not individual occasion use.

Peirce not here talking about this sort of distinction. I put it in for completeness.

Ex. (b) Suppose I find on p. 10 of book sentence “He was a widower” + and on p. 40 sentence “he had lost his wife”. I think it would be not incorrect to say that the sentence “he was a widower” (used on p. 10) + sentence “he had lost his wife” (used on p. 40) had have the same meaning (mean the same); or perhaps to make things quite clear, “the sentences as such, have the same meaning”.

Nevertheless if there is some question whether the “he” in each case is the same person, + it turns out different person meant, I think we might well say “The author didn’t here (i.e. in the two places) mean the same by them because he was referring to different people. It’s clear that when in conversation we get at cross purposes because of muddles about reference of referring words, we can start to clear up by saying,

"Oh but we didn’t mean the same thing …”

I think this sort of things Peirce has in mind; and I think it not incorrect to say that where people use (type) sentences which have the same meaning, neverthelessthey may mean different things (because of difference of reference); i.e. they com what they commit themselves to, and the information the sentence used to convey, may differ in different cases of use of 2 (type) sentences, even tho’ sentences (as such) have same meaning; and some type sentences (unambiguous) may be used to convey different info in different occasions.

But there is another side to medal equally obvious when pointed out, but perhaps even more overlooked; that is that sentences with different meanings may be used to convey the same information.

E.g.IfI say ‘I have a pain’ someone else to convey same information has to say ‘he has a pain’; and no one would want to deny that that [sic] these sentences (as such) have different meaning.

Things having meaning:

a. Things with meaning independent on communication (possibilities) (things with ‘natural’ meaning) [All ‘information’ is a sign of ‘dicent’] 1) cases where x means y and connection purely regulative inferential natural concomitance (e.g. smoke/fire) 1a) likeness between x + y absent (e.g. smoke/fire) 1b) likeness between x + y present (e.g. photograph) 2) cases where x means y and concomitance not purely natural but dependent on human decisions, practice, convention (e.g. tie knots/ wedding b. Things with meaning dependent on communication possibilities 1) ‘rhemes’ (words etc.) incomplete utterances (e.g. ‘rhematic’) 1a) descriptive 1b) logical 1c) referring/indicating 2) complete utterances; 2a) those with conventional meaning 2aa) linguistic 2ab) non-linguistic (e.g. diagrammes, etc.) 2b) those with non-conventional meaning This of course not meant to be exhaustive, inevitable, accurate, or clear

---
Alt.: Reasons for dissatisfaction with interpretants merely as signs with further interpretants which are signs (1) Imprisonment within language; (2) Circularity (no sign [of it] in Peirce); (3) Inadequacy of actual interpretantsto give as adequate full definitions rather than merely working elucidations. (4) non-matching parenthesis (general signs (i.e. those for intellectual concepts) must have (in the end) a general interpretant (a general meaning). What is habit.
---

You may think this trivial and obvious; but often overlooked.

Cf. A. E. Duncan Jones on ‘Fugitive propositions’ in Analysis recently.

And has I think a number of less obvious applications of considerable interest.

Tho’ this not arising directly from Peirce having got so far I think I must briefly show why I said ‘What one commits oneself to by using s’ and ‘information s used to convey’ not identical. Take very obvious case, that will “Smith has left off beating his wife” (s). Suppose this said to me, and suppose I thought know Smith had never stopped beaten his wife at all.

Would I say that what speaker said false (obviously not true). I don’t think I would reply “that’s not true”; I would avoid giving a direct answer and say e.g. “The question whether he has left off doesn’t arise, since he never started.” Have we then a case to which law of excluded middle doesn’t apply? Depends to what he applies Law of Excluded Middle.

In this case I suggest the reason why we don’t apply true or false to statement in question is because can be set out as follows: the information which speaker used s to convey was about the process consisting in Smith’s beating his wife, and was to the effect that this was no longer going on.

But since this process was presupposed That there was such a process was presupposed (implied, IMPLICATED -- Cfr. 'implicature' and 'disimplicature' and 'inplicatura' [sic] as used by Sidonius -- "Latin Dictionary," Lewis/Short), taken as granted), and was not part of information to be conveyed. Nevertheless speaker committed himself to there have been such a process (i.e. he couldn’t have said “Smith has left off beating his wife, but he didn’t start”).

Distinguish then what speaker committed to (i.e. that Smith had beaten his wife, but no longer doing so) from information s used to convey (i.e. about the said process + so, that it [is] no longer going on).

If we want to applying preserve L[aw] of E[xcluded] M[iddle] then it would seem that the obvious thing to do is to say not that it states not that info. Sentence used to convey must be true or false, but that what speaker committed to must be true or false. Suggest we cast great doubt about use of ‘proposition’? ? Sentence express some proposition (1) when has some meaning or (2) when user committed to same thing or (3) when used to give some information? Icon, Index and Symbol

4) Peirce’s next remaining trick is the division into Indices, Ikons and Symbols. Very roughly we might distinguish these as “signs in virtue of causal (temporal or spatial or temporal (not mentioned by Peirce)) connection with an object (Indices)” “signs in virtue of lightness likeness to an object” (Ikons) “signs in virtue of a convention” (Symbols). Peirce says that this division is in respect of the relation of a sign to its object but it also seems to serve as an elucidation of the expression“in some respect or capacity” + ground of representation?

There is evidence that not only ‘mean’, but also ‘sign of’ ‘sign that’ ‘indicates’ etc. seem appropriate in ordinary sense to some of Peirce’s cases examples of indices.

Indices important distinguishing features seem to be “an index is a sign or representation which refers to an object not so much because of any similarity or analogy with it, nor because it is associated with general characters which that object happens to possess, as because it is in dynamical (including spatial) connection both with the individual object, on the one hand, and with the senses or memory of the person for whom it serves as a sign on the other hand … while demonstrative and personal “Indices may be distinguished from other signs by or representations by three characteristic marks: first they have no significant resemblance to their objects; second they refer to individuals, single collections of units or single continua; third they direct the attention to their objects by blind compulsion.

“An index is a sign which would, at once, lose the characters which makes it a sign if the its object were removed, but would not lose that character if there were no interpretant. Such for instance is a piece of mould with a bullet hole in it as sign of a shot; for without the shot there would have been no hole; but there is a hole there, whether anybody has the sense to attribute it to a shot or not.”37 “a genuine Index and its Object must be existent individuals”38 Salient features of indexes seem to be (1) it would lose its charact Both the index and its object are individuals. (2)

Index would lose character which makes it a sign were there no object, but not were there no interpretant [here we can fairly clearly substitute ‘interpreter’]

(3) Refers to object not so much because of similarity or analogy (contrast Icons), nor by asseveration (? = “just makes me think is so-and-so”: “I always associate you with that symphony (because we heard it together), but because of dynamical (causal or spatial?) connetion with object (4) Directs attention to object, and does so by ‘blind compulsion’ (5) Assert nothing (connection with imperative mood) Examples Seaman rolling gait (index of sailor possibly indicates that man is a sailor); weathercock Sundial or clock indicating time of day Letters vici names. a said to b that c’s etc. Yardstick shows a yard (not icon) (seems doubtful case too) Demonstrative pronouns and personal pronouns Words connected with “quantified variables”: any; all; no; some; something; a, a few; nearly all ‘allied’ ‘a few’; ‘nearly all’ etc

Property Both index and object are individuals. (a) Peirce appears to go against this so far as concerns index and when he classifies later demonstrative pronouns as indexical l l not underlined? egisigns. In another place he classifies personal and demonstrative pronouns as sub-indices, describing them as “signs which are rendered such” principally by an actual connection with their objects” and says they are not indices because not individuals. He seems to use Index so that nothing but an individual can be an index, but non-individuals can be ‘indexical’.

Point is of course one can give some rule general rules for the use of a word like ‘this’, or ‘I’; but these would be of course rules governing the use of them in particular concepts contexts, conditions governed their having a reference in such contexts. Moreover tho’ we can give general explanation of some of their use; we should be very chary in general of saying e.g. ‘I’ or this mean such + such. [Translation ‘here’ means ‘this’] perhaps single. In general if asked what ‘this’ means we should turn the question (it is a demonstrative pronoun .......) nb Use like legisigns (type) general statements like Black clouds are a sign of rain (mean rain)

As regards object being an individual, seems to depend on how we interpret ‘object’ It seems to me clear that whenever we use a word like Consider for moment such indices as particular black clouds statements like Dark clouds mean rain Black clouds mean rain (Those clouds mean rain) Those spots mean measles That smoke indicates/means a fire (the smoke indicates a fire somewhere)

There were signs of mice + related All these end with substantives + all can be replaced by ‘that clauses’, and I cannot think of any examples (from the field of natural signs) where this can’t be done.

If ‘object’ being an individual is held to support this can’t be done in some cases, then wrong say object as individual. That is not Moreover we couldn’t say “That fire is what the smoke indicated (after finding it).

I am now putting out what that smoke indicated

However Perhaps what we can learn from black clouds etc. is always no doubt something about some specific individual, or set of individuals But do we not have to know the notion of Similar point seems to apply as regard some fact (e.g. the fact that … can be put in, at any rate in singular ‘mean’ etc. statements). Perhaps then it is to be a defining characteristic of this type at any rate of index this statement about its ‘meaning’ etc. … appropriate sense must be expandable of both asregardssubject and predicate into that clauseswith some spatio-temporal restriction (to distinguish them from cases like

“The fact that How then This would however seem to abut an object or the same that item that class to abut some individual or set of individuals (more or less indefinitely)

Consider together (a) “refers to object because of dynamical connection with object” (b) “would lose character which makes it sign were there no object but not were there no interpretant” (c) “directs attention to object by blind compulsion”

Now it seems fairly clear to me that in connection with some of Peirce’s examples these descriptions read very oddly indeed (e.g. demonstratives) but in connection with others not so oddly. Let’s consider at present the more favourable cases (e.g. what would obviously count as natural signs. What first is force of “refers to object because of dynamical connection” “would lose character which makes it sign if no object”.

Dynamical connection must obviously [be] causal connection, which may (but only may) include spatial connection in any very obvious sense (e.g. exceptional sounds from weather may be indices (mean/indicate/be signs of/telling about) namely indices of behaviour of wood cuttings). In[dex] could have spatial but no causal connection (e.g. “landmark” which would I think fit under the head of indices) Let’s consider only causal connection.

Peirce’s remarks about bullet hole (which means that a bullet has passed through, or was a sign of a bullet having passed throu’) unfortunate.

For obviously there might have been a hole without a bullet (someone might have bred it Let’s consider Suggest what Peirce is saying is then (a) that for a sign to be index (a) it is logically necessary both that its object should exist + that there should be causal connection between sign and object (b) presence of sign of certain such + such kind thus causally connected with object causally necessary for production of the effect on an interpreter which constitutes being interpreted etc.

That is for position of weather cock to mean be index (to be index to someone) of direction of wind (a) the wind must (by itself) be in required direction + (b) being so must be causally connected with portion of weathercock. (b) presence of weather cock linked in this way with direction of wind causally necessary required (meaning) for production of effect by sign on interpreter (being inter[preted] as sign) (what follows of course strenghtens the causal necessity bond about blind compulsion adds something to (b); something more than mere not just that weather cock + causal link known to be there; it looks as if this is all that has to be there, or at least that something else is not required (e.g. contribution from interpreter))

As far as (b) clearly that wind should actually be as indicated not causally required for production of effect on interpreter: most that can be required is general causal connection of weather cock with wind (and interpreter’s knowledge of these; and interpreter might obviously be mistaken about general causal connection, think there was one when there wasn’t).

Of course wind must be as indicated if interpretation is to be correct. That all these causally What causally required then is presence of weather cock plus interpreter belief in general causal connection. What about (a). Suppose we interpret ‘Index’ in terms of mean.

Then Peirce saying to be understood as saying that (in appropriate "sense" of "mean") The position of weather cock meant that wind ne entails (a) wind ne (b) causal connection holds between wind + weather cock. This seems right + useful. For compare different cases of ‘mean’ [conversation at bus stop as bus goes a Those three rings of the bell meant that the bus was full b Was it full?

This seems all right for ‘means’, but doesn’t seem to hold for the other words in terms of which possible to interpret index (e.g. Peirce’s own word ‘indicate’ + allied words) I can say I certainly can say “The position of the weather cock was an indication that the wind was ne, but actually it was se” I think (tho’ not at all happy) I can say “The position ___ indicated that ___ but actually it wasn’t” I don’t think I can say (tho’ again doubtful)

“The position ___ was a sign that ___ but actually it wasn’t” But case obviously clear for that “x means p” (in this use) entails p. For other words analogous entailment doesn’t hold or remain uncertain What does ‘directs attention by blind compulsion’ / ‘would not lose character if not interpreted’ add? What Peirce has Does this tell us anything to distinguish natural signs from others? First not true that all sign which we should count to classify with weather cock (i.e. nat[ural] sign) direct by blind compulsion if this means that that interpreter doesn’t have to think in order to interpret; sign may be “hard to read”.

And conventional signs (symbols) which he wantsto distinguish in virtue of absence of blind compulsion to say don’t work by blind compulsion may be very easy to read.

I think he is in right [on] his part, in that it is a feature which distinguishes “non-natural meaning” from “natural meaning”, that a certain sort of special contribution is required for interpreter, but to say more would be to anticipate later discussion.

Difficulty for his general triadic view of meaning supported by his remarks about character of sign not lost if no interpreter. In sense of ‘mean’ appropriate to weather cock, the it doesn’t see that ‘x means that p’ requires to be expanded to any form involving “means to some one”. It doe It does I think entail something about “meaning to someone”, but only in very timid way. “x means p” seems to entail “if x meant to someone that p, then he would be correct” “if any one interpreted x to mean p then he would be correct”

But this is like: “it is raining” entails “if any one believes it is raining he is right” which obviously doesn’t help to explain meaning of ‘it is raining’. Icons

Icons main features (1) refers to object that it denotes merely by virtue of characters of its own which it possesses just the same whether or not object exists. Anything is an icon of anything, in so far as it is like that thing + used as a sign of it (what does last clause mean?)39 Icon sign which would possess character which renders it significant even though object had no existence (2) [Reference to “algebraical icons”] A great distinguishing property of the icon is that by direct observation of it other truths concerning its object can be discovered than those which suffice to determine its construction40 [e.g. maps would be example; you can plot places by being at distance from given place and the measure e.g. distances between them] (3)

This of importance to Peirce’s theory of inference; but that we can’t go into] Every symbol replaceable by icons (? image or picture: verifiability business) Examples Pictures Architect or artists drawing (for work) Donkey (?) icon of zebras (if I agree by analogy from resemblance of zebras to donkeys, that zebras likely to be obstinate) Maps—diagrams Mental images Mathematical physical formulae (!) Icons distinguished into ‘images’ (simple likeness) ‘diagrams’ (analogy of relations) metaphors (I think this meant to cover donkey/zebra case) nb formulae don’t really conform to test without further benefit; I don’t derive new truths from formulae by inspecting them, but by deriving things from them according to rules. But suspect much of what Peirce has to say about nature of inferential processes (or diagrams) of interest. nb Photograph not icon (because of “physical forcing” of conformity to original) Symbols Symbol (1) Great emphasis on symbols as of conventional nature (2) Symbols must be general in meaning (? because general in nature-type) (3) “Symbols connected with object by virtue of idea of symbol-using mind without which no such connection would exist”41 “Symbol ‘man’ general mode of succession of three sounds or representamens of sounds which becomes sign only in fact that a habit or acquired law will cause replicas of it to be interpreted as meaning a man or men.”42 “Symbol is a sign which would lose the character that renders it a sign if there were no interpretant. Such is any utterance of speech which signifies what it does only by virtue of being understood to have that signification”43 Symbols words + sentences (type) Sometimes combination of icon + index [p. 24] Symbol (a) would lose the character which renders it sign if no interpretant (b) e.g. speech utterance, which signifies what it does only by virtue of being understood to have that signification (c) generalmode (case of word ‘man’ in this case sentence (the use of) general mode of succession of 3 sounds or representamens of sounds [writing] which becomes a sign only in the fact that a habit, or acquired law, will cause replicas of it to be interpreted as meaning a man or men. nb Obvious contrast here with icons + indices, most notably indices (blind compulsion) Again (cf. index) does render = logically ‘makes’ it sign causally ‘makes’ (1)

If ‘logically makes’ then Peirce is saying “logically impossible for p to be sign (symbol) if no interpretant” Cf. in case of speech utterance, when it looks as if Peirce says “speech utterance cannot signify (to someone) without being understood to signify it.” But this looks as if it only amounts to saying “a symbol cannot be interpreted without being interpreted”. True, but of course not peculiar to symbols. (2) Is he saying that peculiarity of symbols is that for a token symbol to be interpreted it is causally necessary that the interpreter should have know what a general prop[osition] to effect that type symbol means such + such? But again this doesn’t seem peculiar to symbols; surely to interpret a particular occurrence of sentence, it is at least causally (perhaps logically) requisite that interpreter should know something general about what sentence (or sentence of such + such kind in such cases) means or probably means. No good trying to avoid by saying that to interpret token symbols interpreter has to know what type symbol in general means (in sen[se] of ‘mean’ appropriate to symbols). This may be true but obviously no use for distinguishing symbols (+ with it special sense of ‘mean’ attached to symbols. Let’s Let means be abbreviation of this use.

Then Peirce would be saying “x is symbol (i.e. means something) if and only if interpretation (= knowing what x means (means ) in particular occasions) is dependent on knowing what interpreter’s knowing what x meanss (in general).” (3) Quotation (1) seems to suggest that distinguishing peculiarity of symbols (i.e. something which meanss ) is that interpretation of occurrence (token) of x depends on interpreter’s having some habit [This not unlike what Stevenson says “elaborate process of conditioning”]. But seems in fact impossibly crude (probably much cruder than Peirce thought) because of his assumption that interpreter’s condition[ing] irrelevant to functioning of index)

(4) He might mean something like this (tho’ no very clear support in text) s is symbol if, for s on given occasion to produce whatever reaction it is which constituted interpretation of s (being taken to mean this or that), it is necessary that interpreter should know or believe that s (type) in general would produce reaction of that kind in people. e.g. to take rocket to mean distress, I must know or believe that that’s how it would usually be taken. (? Not so in case of smoke means fire.) But tho’ this a fair line for getting at important distinction in uses of mean, it can’t do as I stated. You might have ultra-timid race which would never take smoke to mean fire when assured that others so look it; this would not mean that smoke has different sort of ‘meaning’ for them (I don’t think so).

Comments on the Classification

What are salient features of Peirce’s classifications which excite comment:

(1) Demonstratives etc. accompanying e.g. weathercocks under ‘indices’ heading. Admittedly different in that first dicent rhematic + second ‘dicent’, but still queer (2) Icons said be non-assertive non dicent (rhemata) (3) Symbols exclude (a) demonstratives, (b) diagrams + icons generally, yet icon + index (picture + gesture, picture + name) in one place allowed to constitute symbols (+ presumably also to be dicent) (Symbols have to be ‘general’ in meaning) (4) Icons include portraits, artist’s ‘cartoons’, maps + diagrams (as well as logical icons + analogy cases) (5) Distinguishing feature of icons said to be that “by direct observation of it other truths concerning its object can be discovered than those which suffice to determine its construction”

(6) Rhemas include pictures (+ [illeg.] cases of rhematic indices) + words, i.e., things whose sign character status nothing to do with occurrence in dicent signs together with things whose sign character status dependant on occurrence in dicent signs. (7) It seems clear that nearly all the oddities of Peirce’s classification spring mainly from his trying to make single headings do more than one classificatory job at a time (together with one or two rather perverse assimilations). Further that one very important (I think the fundamental) classification of signs (+ distinction of ‘meaning’ sense) has been gained by the classification, tho’ Peirce was not unaware of it, as emerges from time to time in his comments on his classification; that is the distinction between (a) signs which owe their having meaning to their (roughly) their constituting (or capacity to constitute) a communication [or to their function in what constitutes a communication] and (b) signs whose meaning is not a matter of their communicative function (e.g. ‘natural’ signs).

Tho’ Peirce is aware of this distinction, he doesn’t bring it to the fore, + consequently it is never investigated, + it mixed up (in text I mean, not ‘confused’) with other distinctions which seem to be posterior to it from point of view of explanatory procedure, e.g. distinction between conventional/ non-conventional signs, + distinction between linguistic + nonlinguistic signs, + distinction between descriptive and referring (indicating) signs. Take ‘Symbol’ to begin with. Peirce’s accounts of meaning of symbols [e.g. “conventional signs, or one depending upon habit, inborn or acquired”] seems either attempt to confine symbols to conventional signs (with not very satisfactory explanation of conventional signs) or else extend meaning of symbols to such extent as to consider pretty well any sign [depends on meaning of his ‘or’]. Anyway, emphasis on ‘conventional’ nature of symbols pretty heavy. Then it is recognized that e.g. diagrams have highly conventional character; but nevertheless not counted symbols because “we can get more information by direct observation”.

This seems not a bad attempt to distinguish certain conventional or semi-conventional non-linguistic notational devices from linguistic ones (also conventional).

Again, we find it recognized that an icon + index (Peirce’s example bad, but pers. picture + pointing would do) recognized as symbol. Here we do seem to have recognition of distinction between symbol (= here roughly device for communication, whether conventional or not) + other signs. Perhaps this distinction would have been recognized more clearly, if Peirce hadn’t assumed that icon per se cannot be dicent. His motives good (namely recognition that for information to be conveyed we must must have in addition to device forsaying reveal (?) what sort of information to be conveyed, also means for locating or identifying that to which information refers); but he failed to see that icon + sentence may be in just same position in that respect. E.g. suppose I have on wall (a) picture of Smith (b) placard with “Smith is there” written on it. Suppose then loud noise outside door. I could then convey information by taking up and showing to ‘audience’ either placard or picture (with or without pointing). Point is that icon + context alone + non-linguistic context alone, as also sentence + non-linguistic context alone, may “make assertion”, as if were. Similar point in connection with ‘dicent’. Peirce rightly in case of demonstratives,wrongly in case of natural signs connects indiceswith“quasi-imperatives”. This I think why he says weather-cock ‘dicent’.

But it is dicent not because it “as it were tells us to” but because it “as it were tells us that”; i.e. we learn that from it. (This I think of same value as pointer to why one word ‘mean’ has such variety of different uses in connection with “what we can learn from things”). But obviously vast difference between why and which natural signs and other signs tell us things.

This marked by Peirce only by whether dicisign index or symbol. (Implying need for widest case of ‘symbol’, to distinguish communication from non-communication. Similar point already observed in connection with rhemes; rhemes sometimes “elements in” communication, sometimes any non-informative sign. Further complication through tying up symbol with ‘descriptive’ sign (or as he much less clearly says, signs with ‘general’ meaning). This cuts out demonstratives from symbols tho’ obviously conventional; instead Peirce makes a hughly shaky assimilation between demonstratives + indexical signs (which leads him to say very wrong dubious things about demonstratives, e.g. that there is a ‘real’ connection between demonstrative and object. He has (a) failed to realize consistently (tho’ on occasion seems to get across) that tho’ ‘this’ [is] non-general (if that means non-descriptive), nonetheless ‘this’ has general meaning in sense that there are general rules (non defining) for use of ‘this’ (b) failed to notice that from natural sign one learns both what to expect and (often) where to expect it (i.e. indexical sign in a way analogous to combination of demonstrative + descriptive), and has been over imposed by contr very clearly explicit comparison between signs which have function of indicating individual, + signs from which we can learn something about an individual (indexical sign)

Finally class of icons rather a jumble: includ seems to include things that remind one of something (pictures on wall: no genuine use of ‘mean’ applicable here), sketches + diagrams to aid construction of something (no ‘meaning’ here either), use of pictures (or imitative gestures?) in communication (but here often we want to rule out what Peirce describes as distinguishing feature of icons, namely that by direct inspection we can learn more about object than is necessary for construction of icons; e.g. if I draw bird impart information about presence of some bird or other, I shall make drawing as sketchy as is compatible with recognition as bird, just to prevent audience from studying picture further to learn details about bird; I try to avoid making it like a map). This can’t distinguish it from maps + diagrams, about which Peirce good (apart from logical diagrams). Here is point that I resort to drawings + to diagrams if language breaks down; Nevertheless but to drawings, if e.g. no intelligible language to hand, or physical impediment to use of language, to images e.g. if language would be impossibly lengthy or difficult to follow. It doesn’t seem as if mere likeness is ever per se justification of use of ‘meaning’. E.g. photograph (indical) requires causal link; use of drawing or map requires additional special coincidences to make us use a communication willing to say that we have here case of communication.

Let us sum up.

We shall not really be in a position to assess either Peirce’s or anyone else’s general theory of signs untill we have some a fairly good idea of the nature of the distinctions that have been or might be drawn within the body of“signs”; for until we have a fairly clear notion of the similarities and differences among the entities which Peirce would want to call ‘signs’, we cannot really judge down whether a single ‘general theory of sign’ is a possibility, and certainly cannot be in any very good position to decide what question or questions should be answered by such a theory. I shall therefore, to begin with, at any rate, say only as much about Peirce’s views about signs in general as is necessary for as a preliminary to considering his more detailed the details of his classifications of ‘signs’; andI shall try at this stage at any rate to evade the many difficulties raised by what he has to say about signs (‘sign-situations’ in Ogden and Richards’ phrase44) in general.

“A sign, orrepresentamen, is something which stands to somebody for something in some respect or capacity. It addresses somebody, that is, it creates in the mind of that person an equivalent sign, or perhaps a more developed sign. That sign which it creates I call the interpretant of the first sign. The sign stands for something, its object. It stands for that object, not in all respects, but in reference to a sort of idea, which I have sometimes called the ground of the representamen” (Buchler sel. p. 99)45 He goes on to give an extremely unclarifying explanation of his use of the word ‘idea’ here, which fortunately I think we ignore. Sense can be made of the notion of the ‘ground of a representamen’ if (I hope the right sense) if the word ‘feature’ or ‘property’ or ‘characteristic’ be substituted for ‘idea’. He is I think simply saying that if something x ‘stands for’ (‘is a sign of’ ‘represent’) something y, this is always because of some particular feature which x has (e.g. its likeness to y). The sign-object-interpretant is, he says, “a genuine triadic relation, that is its three members” (or better, terms)“are bounded together by it in a way that does not consist in any complexus of dyadic relations”46 “A sign is a representamen with a mental interpretant”47 cf. beginning of 1st quotation (in the mind). In the passage from which these quotations are taken, and I think in many other places in Peirce’s work, the an interpretant of a sign is regarded as being ‘mental’ in the sense of being a ‘thought’ (tho’ this is complicated by his occasional distinctions between different kinds of interpretants; however this does not affect the fact that the kind of interpretant with which he normally concerns himself is something which he would normally describe as a ‘thought’). But it is Peirce’s view that thought itself consists of signs [cf. Buchler sel. p. 233 “The third principle whose consequences we have to deduce is that, whenever we think, we have present to the consciousness some feeling, image, conception, or other representation, which serves as a sign.”48 The development of this view lands leads him into some of his murkier obscurities, but fortunately we need not, for the present purpose, go into them.] This view that thought consists in signs clearly lands Peirce with an infinite process: for if a sign s1 has to have an interpretant s2 and s2 is itself a sign, then s2 has itself an interpretant s3, a sign which itself … etc. This consequence of his theory of signs Peirce himself points out in the passage under consideration, and (in the passage) shows no sign of unwillingness to accept it; in fact he regards it as involved in the “familiar idea of a Sign.”49 Elsewhere (in a later and important article) he finds it necessary to say more—and on somewhat different lines— on this subject. But on that anon. Salient points so far

(1) A sign stands to somebody for something (in some respect). This involves a signs being a actually (or possibly) related by a triadic relation to an object and an interpretant

(2) This triadic relation is ‘genuine’

(3) The interpretant is a sign thought + so a sign, and therefore an actual or possible sign is a member of a further triad, which generates yet a further triad and so on ad infinitum. [p. 30] Now the following seem to me to be some of the more pressing questions which arise out of the passages I have just quoted or summarized. (i) What is a theory of signs a theory of? Are we to understand the word ‘sign’ in its normal every day sense, or is it here used in a non-everyday, technical way; if the latter to what purpose? (ii) What is the force of Peirce’s contention that the ‘sign-relation’ is genuinely triadic?

Does this rule out further explanation of the nature of the ‘relation’? If not what sort of explanation would Peirce favour?

(iii) Is the infinite process series of interpretants of any sign tolerable?

Does Peirce consistently tolerate it? These questions cannot be dealt with without some further discussion of the question what an interpretant is, and this will I think be found to have some bearing on the questions under (ii) (iv) What about the ‘object’ of a sign? Does such a notion involve hypostatization? About these questions I shall at this stage anyway say as little as possible. i. What is a theory of signs a theory of?

I think we can answer this question (as regards Peirce) straight away by saying that when Peirce says something about signs in general, what he says can be understood as saying something in general about ‘meaning’ (a) weak support for this contention comes from noticing the words he uses when talking about (+ applies to) ‘signs’ in general, e.g. ‘stand for’ ‘rep ‘represent’ (as in representamen), and elsewhere ‘denote’. All these words (and sign itself) are traditional encounters in philosophical discussions of meaning.More (b) (I think quite improperly used encounters) (b) More important cf. see Buchler Sel. pp. 91–92.

Here he maintains that meaning is not only a genuine triadic relation, but also is involved in any genuine triadic relation. In support of this latter (not very intelligible) contention he considers examples of triadic relations and claims that the idea of ‘sign’ (as defined in the passage I quoted earlier) is involved (e.g. in the notion of a, b, c being identical).

It is clear that for him ‘having a meaning’ and ‘being a sign’ are interchangeable expressions. (c) Finally let us take more or less at random some of the things that Peirce would wants to classify as ‘signs’; the rolling gait (of a sailor), a diagram, the a shout of “Hi”, the word “that”, the any sentence (which is not nonsense), a weather-cock. Now what for most of these there seem no great difficulty in framing sentences containing word “mean” which I am fairly surewould be able to replace sentences containing the word“sign”which Peirce [would] be ready to employ about the things in question. e.g.

(1) “a rolling gait means (probably means) that the its owner is a sailor” // “a rolling gait is a sign of a sailor”

(2) a “that one is the ripest” b “which one do you mean by ‘that’ (‘that one’)” [Not very natural, but cumbrous rather than contrary to usage]

(3) “The weather-cock pointing south means that the wind is in that direction”.

The only one about which there seems some difficulty is the diagram (he doesn’t naturally speak of diagrams as ‘meaning’ this or that); but I think one could (tho’ alternative locutions would usually perhaps be clearer and come more trippingly to the tongue) talk about what parts or features of a diagram (or map) mean: e.g. if, considering a Mercator’s projection map, I find that the shape marked ‘Greenland’ is approximately the same size as the shape marked ‘Arabia’ [I’ve no idea whether or not this is in fact the case], I may ask (or answer) the question “Does the approximate equality of size of that and that mean that Greenland and Arabia are (really) about the size same size”; or looking at a diagram of some kind of machine and noticing (say) some curious (nb) parallel sort of parallel lines circles, I might ask “what do these mean?” or“do these mean anything?”What seems me to emerge as a suggestion from the isolated examples in clauses, is that it may be natural to use ‘mean’ (a) when one knows or thinks or suspects that some ‘distortion’ is involved (as in case of ‘Greenland’ etc.) (b) when one knows thinks or suspects that features in question have not a merely representative character, but are in some sense conventional (which is what I had in mind in second case (‘circles’). It is important to remember that maps and diagrams often or always embody ‘conventional’ features or elements. One thing I think is quite certain; when we should never point to an area on the map and say “That means Arabia”; and I am afraid that Peirce would want to give as answer to question “what is map or diagram sign of”, “that of which it is a map or diagram.”

This if correct means of course that to explain ‘sign’ in terms of meaning will not wholly fit Peirce’s use of ‘sign’. But I think the substitution will work fairly well; in most cases the answers to the question “what does x mean” would correspond fairly closely with the sort of answers Peirce would give or might be compelled to give to question “what is x sign of?” Moreover I feel fairly confident that no other word but “mean” would do anything like as well. It is here important to note that“sign” in any ordinary, un-technical sense cannot be applied to a considerable number of Peirce’s examples of ‘signs’ [e.g. ‘this’, ‘Hi’, a sentence, if we consider merely above-quoted examples].

The same applies to other words verbs which occur in Peirce’s explanation of sign—‘stand for’, ‘represent [representamen]’. [p. 31] I recognize I’ve made quite a bit of the question about the relation between “meaning” and “sign on the one hand and Peirce’s words (“sign” etc) on the other, but I make no apology for doing so. For in general the use (unannounced, [un]heralded) of technical or crypto-technical terms leads to nothing but trouble, obscuring proper questions and raising improper ones. In this particular instance good reasons for discussing and emphasizing the need to restate Peirce’s remarks in terms of “meaning”.

(a) Peirce’s principal word “sign”, when it occurs in contexts in which there is some reference to a person, does so most naturally most clearly in contexts where something is a sign to someone (or ‘for’ so. That is to say in contexts where the person concerned is roughly speaking in position of audience. (I don’t of course say it can’t be used otherwise). Mor If we do think of persons concerned thus, we may tend to neglect fact that some at least of entities Peirce calls ‘signs’ can be used as well as ‘encountered’, and that this may be very important. It is clear that Peirce himself usually tho’ not always does tend to think of persons as ‘audiences’; and this no doubt has some connection with people who do this tend to hold a causal theory of meaning, or at any rate persons who hold a causal theory of mean[ing] tend (in the first instance at least) to treat persons as audiences. Important in view of question about Peirce’s own position vis a vis causal theory.

(b) Some of words in Peirce (‘stand for’, ‘represent’) seem to me reflect, help, or give rise to, inclination to think of the meaning e.g. of a word as always an object (if not a thing, something at least which is a particular, or something at any rate which is for philosophical purposes not too unlike a thing). “This picture represents the death of Napoleon (or Napoleon in his youth)” “This match represents the bowler Lindwall etc.” “h. stands for ‘Henry’” “[nb ‘stand for’ ‘represent’ much less common than ‘mean’: philosopher’s use this with unfamiliar subjects (grammatical), which “Represent’, + ‘stand for’ also seem to be connected with notion of ‘substitution’; and tho’ this may be helpful, if generalized (as the use of verbs in questions perhaps induces us unconsciously to do) may reflect, help, or give rise to the suggestion that for every word there is something for which it is a substitute. n.b. Russell’s use of ‘stand for’ in discussion of universals (145–146) in Problems of Philosophy.51 ii. Having I hope justified my intention of treating Peirce as attempting to give some sort of theory of meaning, I want now to ask what it is. I should say in advance that I do not think it possible to find any very detailed analysis of meaning in Peirce: he was too volatile and inconsistent for that; but we may hope to find motives of a theory (or theories), which could be useful as introduction to working out of at least some of these by others. First can we re-state in terms of ‘mean’ Peirce’s view that ‘sign-relation’ is triadic? I take it this would have to claim roughly that wherever we get a sentence statement e.g. beginning “x means ....” or “x meant ....” this would always be capable of expansion into, or analysis in terms of a statement such as “x means (meant) ..... to ...... (person or persons)”. Now as regards these there are one or two things he clearly does not mean

(a) He very obviously does not mean that any statement about x meaning something entails that x has meant or will mean some is actually understood (interpreted) by any particular person at any particular time. E.g. it might be true that a recherché or difficult sentence might mean something, without in fact ever having been interpreted by anybody or even uttered (here must be some meaning for sentences which have or never will be uttered). Possibility of meaning to, being understood, being interpreted is the essential, not relative interpretation. (b) Since being interpreted by, or understood by, (actually ‘meaning to’) seems to involve the presence of an interpretant which is itself a ‘sign’, either the interpretant sign needs not itself be unde and has a further (possible) interpretant, it works as if either there is an endless series of constantly present interpretants or else we at some stage come to an interpretant ‘sign’ which tho’ perhaps capable of being understood is not in fact understood. But I don’t think Peirce means if asked would say that either of these alternatives is the case. An interpretant ne “It is not necessary that an interpretant should actually exist”. An interpretant “in futuro will suffice”.

I think The regress of natural interpretants should be avoided by his saying that e.g. I can und it might well be true that I understood a remark made to me even if no interpretant ‘sign’ came before my mind, provided that I could provide an interpretant, i.e., provided I would have done do so if some further (unspecified) condition were fulfilled. [I am not sure that Peirce always thinks in this way about the original interpretant in the series, but this has to do with what I have to say about interpretants.]

What then is the habit in question, which constitutes ultimate logical interpretant?

Very obscure here.

Some things clear (1)In some sense or other effect of sign (or of semiosis of sign, whatever being an effect of that is; I suspect this no more than to emphasize triadic relation of meaning, even if causally explicable) (2) Though not identical with experiment/observation role, it [is] obviously fairly closely connected with  (3) Not merely a habit but controlled (deliberatively formed) habit. The conclusion of (formed by?) ‘ideal experimentation’ i.e. imagining oneself in certain circumstances + doing certain things (4) Habit in question described as habit of action in a given way whenever we desire a given kind of result;53 and earlier in article belief said to readiness to act given certain conditions and motives is said to be belief ([in the] same earlier passage conjecture which is ideal experimentation [is] said to ‘be expression of’ belief or equivalent to a habit of kind in question.54

(5)“The most perfect account of a concept that words can convey will consist in description of habit that concept calculated to produce. But how otherwise can a habit be described than by a description of the kind of action to which it gives rise, with specification of conditions + the motive”.55 Seem to be 3 alternatives (1) habit of giving (uttering) experiment /observable result rules (when stimulated by occurrence of sign) (Buchler: says no) (2) habit of making confirmation, testing / of confirming the proposition expressed by a sign (Buchler says yes) (3) habit involved in belief in proposition expressed by sign (Grice +? Gallie) Can’t make real diff[erence] between them. (3) Seems to tie in best with talk of belief + habit; seems however to commit him to holding that sentence must be believed if (fully) understood; but perhaps avoidable if belief needn’t be actual but only would be brought about by sign in certain cases. (1) [and] (2) seem altogether to ignore connection with belief. (3) would be like point of Stevenson; descriptive meaning = roughly tendency to arouse belief. [p.10] This might link with the operation-sensible result rule, since tendency to perform the operations if one desired the result would be one of the tendencies of which the habit (belief) would consist. But I admit very unclear a lot of stuff I’ve not taken into account. If this right he could be holding a causal view of a Stevensonian type mixed up with meaning-explainability theory Confusion are to (1) explainability seems requisite of meaning function, whether or not it will do as def[inition]; it would no longer follow from definition; so you might skip over to it from another theory. (2) (a) ambiguity of interpretation (b) his unexpected saying ([illeg.] of def[inition] of sign) that meaning is not object but interpretant (no doubt reasons for this) Buchler however has 3 suggestions 2 of which I can more or less understand (1) habit is one of giving operation-result type rule when stimulated by [illeg.] sign (in one’s own thought or otherwise?)

(2) habit is one of going about confirming sentence (meaning “meaning of sentence is habit of going about confirming it.56 This [is] important which in (1) [illeg.] as interpretation but (1) would just be expression of belief that final will did give meaning of e.g. this island (2) doesn’t seem very [illeg.] for one to have (if it just means that one would if necessary confirm it doesn’t seem to tie with belief / still Object: usually roughly I think denotation as opposed to connotation; interpretant connected with connotation but still here (1) interpretant said to be representation “representamen of same object as sign” (= has same denotation?) if we suppose interpretant not always synonymous/definition with sign (seems [illeg.] account of anything we can do) then why bring it in at all. Avoiding metaphysical “objects”? But object has to be there anyway by def- [inition] (But this might be got over by his unintelligible distinction between 2 kinds of object tho’ I doubt it)

Interpretants (a) Emotional Interpretant i) feeling = sensation (music conveys + intended to convey composer’s ideas, usually feelings57) ii) feeling of recognition (a feeling we come to interpret as evidence we comprehend the sign:58 cf. immediate interpretant “total unanalysed effect sign calculated to produce: and I have been accustomed to identify this with effect sign first produces on a mind without any reflection on it.59 (b) Energetic interpretant—“functions through mediation of emotional interpretant, and suchfurther effect involves effort” (“ground arms”: en[ergetic] Int[terpretant] muscular effort, but more usually mental effort60 Not meaning of a sign because individual (?) (c) logical interpretant (the ‘meaning’ of sign) “in its primary acceptation translation of sign into another system of signs “the meaning of a sign is the sign it has to be translated into62 (nb confined to “intellectual concepts”) Possible theories (1) causal effects or tends what sign effects or tends to effect (a) behaviouristic (b) non-behaviouristic (2) intention e.g. intention of user e.g. Ewing63 (3) ability to give rules e.g. logical interpretant Buchler admits he speaks in (1) like language of time but thinks he really hadn’t to (a) relation to metaphysics (only applies against behaviouristic alternatives) (b) restriction with words like represent [,] stand for; these he thinks obviously require intention (only in theory do, but can’t see any evidence that this obvious to Peirce) (Question is for Buchler that only [illeg.] ‘represent’ as treating something for certain purposes as it may involve intention but not obviously user’s intention) (c) logical interpretant but what it refers to doesn’t help again (d) things he objects to Lady W[elby] that intention won’t do because natural signs have no utterer64 [maybe means sign special department intentional theory] (e) [illeg.] of interpretants as significate effect [rep. in Buchler 276 seq. “Pragmatism in Retrospect”65] Why does Peirce in later pages getsworried about logical interpretant as sign. Series of thoughts as interpretants must be ended by ultimate-logical-interpretant which (sign) is not a sign (at any rate in same sense as interpretants are) but is a habit. (1) ? imprisoned in language: but why bring in habit: not different kind of rule [This fact is the rule] (2) Circle difficulty: no sign that he is [illeg.] symbol [illeg.] it. But not aware.

Notion of logic ultimate logical interpretant highly obscure.

Some in his theory clear. [p. 28666]

(1) Not purely habit but controlled (dependent on deliberate formation): [illeg.] same nature tho’ not a thought (2) Engendered (“the conclusion of”) mental experimentation “imagining yourself acting in certain ways (3) Trial rules of form (given experiment observation their sensible result) not enough because “doesn’t tell us of the essential nature of effect of the on the interpretation of semiosis of sign” (semiosis explain as action or influence involving sign/object/interpretant. Nevertheless habit in some way connected with first type of rule. (4) Habit described as habit of action given certain conditions and motives; and readiness to act given certain conditions and motives is earlier said to be just belief (same earlier passage conjecture seems to be expressive of or equivalent to a habit of kind in question) Looks then possible that habit might be quasi belief or conjecture (quasi belief) in what sign (not interpretant) asserts (habit of ‘imagining’ oneself acting as if …)