User:Jorge Stolfi/Signs and Symbols

From Wikipedia, the free encyclopedia

This note reorganizes and expands a discussion I had with another user about his/her proposal to merge sign and symbol and treat the two words as synonymous. The proposal was apparently motivated by the previous definiton os semiotics as "the study of signs and sign systems". Each section below was actually a separate response; see Talk:Symbol for the whole exchange.

Round I[edit]

I strongly object to merging the two pages. Sign and symbol are NOT synonyms, although they have SOME senses in common. No one would say that cough is a "symbol" of influenza, or that the Chrismas tree is a "sign" of Christmas. Rather, this article must point out that one of the meanings of "symbol" is a "sign" (specifically, a "grapheme") usually stading for anything other than a sound.

Round II[edit]

It seems that there is here a confusion between REALITY, PHILOSOPHY, and LANGUAGE. Greek English

Language first[edit]

In English the words "sign" and "symbol" cannot be used interchangeably, so they are not synonyms. For one thing, "symbol" is almost always used for things whose meaning is established and retrieved by convention, even when that convention is based on physical association; whereas a "sign" (in the sense of "evidence") is almost always used when the meaning is established and retrieved by logical or physical necessity. Thus cough may be a "symbol" of influenza in a syrup advertisement or a medical brochure, but it is a "sign" to a medical doctor. A wooden crucifix burned in a Satanic ritual is still a "symbol" of Cristianity but hardly a "sign" (evidence) of it.

Conversely one may use "signs" (hand gestures), but not a "symbols", to show the way to the restroom or speak to the deaf; but bowing, although it is a "sign of" respect, is not a "sign" in this sense. The "signs" that are placed along roads or afixed to walls may have "symbols" printed on them, but are not "symbols" themselves. The "sign" of the number 3.1415926... is positive, whereas its "symbol" is π. A yellow spectral line is a "sign" of sodium, but is "symbol" is "Na".

This may not seem logical, but that is how the language *is*. You may wish that the two words were synonymous, but they just aren't. In summary, "Symbol" almost always means "anything that, by convention, can represent something else in information communication or processing"; whereas "sign" has several very distinct senses, one of them being "evidence" (idependent of whether there are humans, computers, or conventions involved), another one is "hand gesture", another one is "which side of zero", yet another one is "non-phonetic glyph"; and only the last one is a (special case of) "symbol". (Except, however, that an unintelligible pen mark may still be a "sign" but not properly a "symbol"...)

Now, semioticists may have decided to redefine the word "sign" to encompass the common sense of "symbol" and perhaps some more. That is their right, just as topologists redefine "space", graph theorists redefine "edge", and chemists redefine "phase". However that redefinition should be confined to the semiotics and sign (semiotics) pages, and has no right to take over the general sign and symbol pages.

As for the philosophy[edit]

You seem to be saying that semioticians expand sign (semiotics) to include not only symbol (things that have conventional meanings to humans) but also sign (evidence), e.g. the syntoms of a disease or the spectral lines of an element, or perhaps even any kind of physical effect whatsoever.

Round III[edit]

It is not the goal of Wikipedia to fix the English language.

Sign and symbol are not synonymous[edit]

As for "signs" that are not "symbols", consider "fever is a sign of infection": no one (except perhaps semioticists, and not even all of them) would agree that "sign" can be replaced by "symbol" here. Consider also "please call the hardware store and order a dozen blank 1' x 1' signs". Sure, the distinction is "artificial", but so is everything that was created by humans - including the English language and semiotics.

The two concepts[edit]

I still sense here a confusion between language (words), philosophy (concepts), and reality (objects and events). Consider these two CONCEPTS:

(1) any entity A which humans (and computers on behalf of humans) arbitrarily

select to represent some other entity B, and which they manipulate in ways that are in some sense analogous to processes that they believe areproper to B;

(2) any phenomenon A that, by natural physical processes, has its probability significantly increased after the occurrence of some other phenomenon B.

These are certainly distinct concepts, one distinction being the arbitrary nature of the association in (1), and its being defined only relative to human action and interest. Said another way, one cannot claim that an entity fits (1) unless there is someone *using* it that way; whereas an entity will or will not fit (2) independently of human involvement. Moreover, an entity A may be, at the same time, associated to B by one person, to C by another person, and unassociated to a third person.

Broader concepts and poorer theories[edit]

Now, of course one can define another concept (3) which is simply "(1) or (2)", but this does not make (1) and (2) the same. Moreover, there are quite a few interesting things that can be said about concept (1) which do not apply to (2), e.g. how humans choose the entity A, how they communicate their choice to other humans, how popel react to discrepant associations by others, how the association between A and B may evolve from a individual choice to a cultural standard, etc. etc.. (This by the way, is the sort of material I expected to find in the semiotics page). Likewise, there are many interesting things that can be said about concept (2): in fact, experimental science is basically the study of (2). Now, the statements that are true for any (3) are much fewer than those that are valid for (1), or those that are valid for (2); this is just simple logic. In fact, I cannot think of any non-trivial statement that would apply to all instances of (3), and I claim that the theory of (3) is practically empty.

Of course there are many entities that belong to both concepts: in fact, when humans have to pick the entity A in (1) they generally find it easier to choose one that also fits (2). But this too does not make (1) and (2) synonymous.

The two words[edit]

Now, there is this English WORD — "symbol" — that, by common usage, has been attached to SOME entities that fit (1). Another WORD — "sign" — has similarly been attached to a SOME entities that fit (1) and a SOME entities that fit (2), as well as to other entities that fit neither concept (such as the "blank signs" above). These attachments are certainly arbitrary, perhaps illogical, but the only way to change them is to convince a majority of all the gazillion English speakers in the world to change the way they use and understand those words. Just because a community (the semioticians) have agreed to attach the word "sign" to a different concept does not authorize us to use that attachment in Wikipedia on a page that is directed to readers in general.

Moreover, just because the meaning of the WORD "sign" spans concepts (1) and (2), it does not follow that those two concepts are identical, or that "sign"should be redefined as meaning (3), or that concept (4) — defined as the set of all entities which can be called "sign" — has any philosophical interest. In short, one should never confuse WORDS with CONCEPTS.

Concepts are not reality[edit]

Finally, one must keep in mind that concepts like (1) and (2) are still artifical constructs of the human mind. There is nothing in the real world that compells us to combine John's pulmonar spasms at 10:35:23 with Mary's pulmonar spasms at 20:12:44 into a single "cough" concept. In fact, there is no neat distinction between between a spasm and normal breathing, and no clean way to tell when spasm begins and ends. In particular, even though concept (2) seems to be defined in terms of natural laws and not convention, it is still an artificial construct because "phenomenon A" is an artificial concept.

It follows that reality is not bound to conform to any conclusions that we may get by reasoning about concepts. In other words, one must be careful not to cofuse CONCEPTS with REALITY.