dabs of grue banner

Domestic and dictionary meanings in natural language - remarks on the delusions of referring and the particularity of proper names

(These remarks are not very well arranged. To get the full picture you might need to read to the end. Naturally I believe this picture is right, more conventional theories wrong, though it's unclear how the conflict might be resolved.)

People refer to things, or believe they do, in almost every conversation. But the fact that they somehow manage to refer to things even when those things don't exist begs the question how they manage to refer to things that do exist. Moreover there are all kinds of entities pervading the zone between the unmistakably real and genuine, and the unmistakably fictitious and fake - a changing landscape becoming ever more dazzling, if not disconcerting, in our media-driven world. It would be absurd to suggest that each kind of creature be lodged in its own quasi-existential cage along with a unique prescription for referring to it. In everyday parlance, at any rate, one size fits all.

The act of apparently referring, and of apparently doing so successfully, of course requires a recipient who understands and accepts the facts presented (normally the truth of an entire statement). Yet most people either cannot always distinguish factual statements from fictional ones, or deliberately choose to immerse themselves in fiction, treating it temporarily as if it were factual (as when reading a novel). Nor are these phenomena restricted to the domain of verbal communication. People react to things that only exist as shadows of the real thing (for example, characters in a movie), and behave as if an entity exists even when it apparently has no shadow, apart from verbal chatter (for example, God). All this seems to be in keeping with the now quite well substantiated theory that what people actually react to and refer to are conditions/models in their own brains, or the corresponding "mental" conditions/models.

The main question to be addressed here is: why do logicians and semanticists treat singular terms, in the context of their usage in propositions, so very differently from general terms, when the only obvious difference between them is one of number? In particular, why do they regard those eminently referring devices, so-called proper names, as a totally different species from descriptions and general nouns? For sure, they would come up with plenty of reasons (because their brains work that way), but my optimistic aim is to banish at least the most important ones. I shall not attempt to deliver a sustained argument (because my brain doesn't work that way), preferring instead to jot down a few remarks in defence of the view that there are no crucial differences between general and singular nouns, aside from indexicals when used demonstratively (that is, words like this, there, now when used as direct pointers to objects, actions or events in the immediate surroundings). Against this, however, I shall maintain that language engages "non-logical" context-related differences of meaning which do indeed often reflect the classical distinction between singular and general expressions. These will be called domestic and dictionary translations (of words, expressions, statements), although in practice the boundary between them is indefinite.

Once having disposed of the traditional referring association between language and real objects, however, the continuing pursuit of a hard link remains imperative, for the scientific notion of truth seems to depend on it. With that prospect in mind, therefore, these notes will end with a segment on statements containing demonstratives and related kinds of statements (anchored propositions - see below), as well as signs that barely count as language at all.

First a couple of terminological formalities: For the sake of simplicity I shall continue to speak of "reference", "referents" and "referring" despite my conviction that language by itself lacks the wherewithal to refer, in the conventional sense of that word. But, in regard to domestic parlance (and borrowing from a nautical vocabulary), I shall also draw a distinction between anchored referring statements - made in the immediate presence of the object or event being referred to - and floating referring statements - made in the absence of the object or event being referred to. Again, while in practice the distinction is not sharp I'll suppose that it is, for reasons that should become evident later on. Almost all of what immediately follows relates to floating statements or their supposedly referential components.

I'll begin with a few blunt dismissals. In any language the usage and meanings of words of all kinds, including names and descriptions, arise out of some historical context or other, usually obscure and evolving but sometimes transparent and instantaneous. However, (1) I can see no good reason for distiguishing the ways in which the meanings of names of individuals arise and the ways in which the meanings of general nouns arise; and (2) the idea that floating reference somehow depends on historical evidence is clearly absurd. Regardless of whether there's an interested recipient, referring in the absence of the object referred to is a purely verbal act that depends only on the knowledge of the referrer. Especially worrying is the claim that the reference of a proper name harks back to a naming event (a "baptism", or something of that sort). In ordinary conversation when a person refers to someone by name nothing could be further from his mind (unless he's a philosopher) than the origin of that person's naming, and no such contrivance can underpin either the intention or the success of the referring event. Even more preposterous is the notion of rigid designators, as applied to both proper names and "natural kinds". Unfortunately the main features of this causative theory of names (the invention of Kripke, Putnam and others in the "externalist" camp) appear to have become widely accepted.

At the other extreme, as it may be, stands the currently unpopular theory of reference prescribed by Meinong, from which I also wish to alienate the present approach. According to Meinong, as I understand him, there's a realm (indeed several realms) of non-existing objects, such as unicorns and the Golden Mountain, which nonetheless have some kind of "being" and can be referred to just as one can refer to real objects, in the traditional sense of "referring". Against this, my view is that in using ordinary language (if not all language) we can't really refer to anything at all, existing or not.

Consider the statement “My sister is in the shopping centre”. If I have a sister and there's a particular shopping centre that "the shopping centre" picks out sufficiently well in the current context and my sister is there, then the statement is true. In other words, according to the conventional wisdom, if all the terms in the statement correctly refer to existing objects or real relations between objects the statement is true. And in the circumstances that it's untrue (and/or "incompetent"), although the statement still makes sense - the same sense or "connotation" it has in the circumstances that it's true - not all its terms have referents. The "domesticity" argument rejects this widely held view.

According to the domesticity proposal the apparent definite descriptions here are achieved by a narrowing of context concomitant with an amplification of meaning, rather than by an application of indefinite meanings to specific objects and events. In other words (and greatly oversimplifying the picture) dictionary meanings are displaced by domestic meanings. But if I have no sister or there is no relevant shopping centre, then the statement fails because the words that imply a narrower context (normally the current domestic context) in fact don't have the richness of meaning expected of them. In that case we might say, not that the components of the statement lack reference, but that they lack those richer meanings, or that they have not acquired any relevant meaning.

So, if I have no sister, the falsity of the proposition “I have a sister” consists not in the lack of reference but in the lack of potential sense of "my sister". The terminology of “I have a sister” must of course be understood in the dictionary sense, but if the proposition is true, it is true because of the additional connotations that can be attributed to it, and not on the ground that there is an x that fits the “bare” meaning the proposition.

Of what are the "additional connotations" of "my sister" composed? Well, supposing I did have a sister (call her Rebecca Robinson, if you like, but we'll consider proper names later on), then surely the phrase “my sister” would then be extremely rich in meaning. It would be answerable to all sorts of further detail, a great deal of which would arguably be part of the meaning of "my sister". And this would be what I would call the domestic meaning of my sister. It is this kind of meaning that is totally absent from the idea of my sister in the circumstances that I do not actually have one.

On debut, one might say, the proposition "I have a sister", while varying according to context, possesses something close to its dictionary or bare meaning – it asserts little more than what the words say, merely that I have a sister, regardless of anything else that might be true of my sister. If I do have a sister the proposition is true, if not, false, and in either case the sense of the proposition appears to be the same. The point in question is what else is involved in saying it's true as opposed to saying it's false. A natural answer is that it implies that my sister denotes a certain person. But the claim of the domesticity theory is that denotation is, so to speak, a window to pure description – and usually very rich description – and not to be passed off into meaningless variables. The kind of description(s), however, matters. It would hardly be good enough (in this instance) to say that my sister is 160 cm tall with curly red hair and blue eyes. The description(s) must be adequate to bring her squarely into the correct domain, the public domain of which I am a member. For the "problem" with people and other living things is not only their extreme intrinsic complexity, but the complexity of their behaviour and interaction with their environment. My sister (let us assume, and according to one of many possible descriptions) is someone you can bump into at the local shopping centre and have a conversation with about the weather, who owns two cats, loves to eat nachos, plays netball on Saturdays and sleeps lying on her back. (Also see Existence: #12 Accommodation)

In everyday conversation, even when you're talking about someone else's sister whom you have never met and about whom you know nothing, it's most unlikely you'd have a dictionary definition in mind. Rather, from the outset you'd be thinking of a person whom your listener calls "Sis" and who answers to questions such as "does she live with you?", "how old is she?", "has she got a dog?"..... In other words, you'd immediately go down the domesticity path; so general noun-expressions of this sort could be called germinal domestic expressions.

Whilst assembling this article I was struck by these interesting statements that were broadcast on TV*:
"Anyone can be a father but it takes a real man to be a dad"
"To the world you were one person, but to our family you were the world"
- bravely and lovingly spoken by the daughter of murdered police officer Brett Forte, at his funeral in Toowoomba, Queensland, 7 June 2017. From a Queenslander's point of view this was the ceremony of the century, poignant, inspirational, and attended by a massive crowd, an impassioned tribute to the entire police force, not just the fine officer who lost his life. Surprisingly both quotations illuminate the dictionary/domestic variation in meaning. In the first statement the word "father" answers to a dictionary interpretation while the word "dad" is a good example of the germinal domestic category. In the first clause of the second statement the word "person" signifies, rather vaguely, a particular person known or unknown, and seems to transmit somewhere between a germinal domestic and a dictionary meaning. In contrast, the intent of the second clause is decisively domestic.

*I hope I won't be compared to those politicians who have no qualms about exploiting events like this for political gain!

It's useful to note the analogy that can be drawn between the dictionary/domestic distinction and the formal probability/prophecy distinction exemplified in the story recounted in Existence: #17 (please read this first). When the seer's version of events is described in words, that is, rendered as propositions, a seeming paradox arises. The formal probabilty of the seer's prophecy being true is greatly reduced and indeed rapidly approaches zero as more detail is included. If our analogy is appropriate, it implies that domestic propositions, considered as conglomerates, are much less likely to be true than dictionary propositions (general definitions). Therefore, far from bringing us closer to referring as construed in a traditional context, domestic parlance appears to head us in exactly the opposite direction. This is because raw definitions convey much less information than domestic descriptions, and presumably propositions of any kind carry less information than the perceptual (etc) data upon which they are based. One might almost say: language moves out when experience moves in. Or, insofar as truth is a characteristic of propositions, and absurd though it sounds: truth ends where experience begins. Another nail in the coffin of accommodation! (Also see #10 below).

Consider the narratives of pure fiction and mythology. Some such domains contain beings that are descriptively very rich. But can we refer to them? The fact is, we do refer to them (or rather, we suppose that we do). Just as in real-world communication, fiction contains the language of reference. Furthermore, if we can be said to refer successfully in real life, we can equally well claim to refer successfully in fictional contexts – no question about it. For language alone cannot guarantee the existence of what it "refers" to. In other words: although the majority of domestic-looking terms in fictional contexts might not refer to anything in the real world, nevertheless they seem to refer, and this can only be through connotation and grammatical structure. Why should we suppose that real life communications are any different? Except in certain circumstances where the referents as such (the actual objects of statements) effectively become included in the statements themselves, the meaning of words is connotational: it is understood in terms of mental images and verbal relationships. That is to say, floating referring statements won't suffice, we at least need anchored referring statements to enable genuine reference, and it could be argued, even more restrictively, that in the absence of demonstratives used in situ, language does not possess the wherewithall to connect with reality.

An important point about these "pure descriptions" is that they must themselves be true in some relevant and adequate sense. And this requirement theoretically leads to an endless regress of dependency on description. In the present example relating to my non-existent sister, the proposition is false and so, therefore, would be most of the propositions I could muster to provide concrete support for her existence. Mere talk is of no value now, just as it’s of no value for the crook giving false evidence in court.

This might lead one to presume that language can never get to grips with more than what I've called the accommodation of a concept; and that even here it faces extreme difficulties. For it seems, firstly, that no amount of description can ever succeed in definitely pinpointing an object and, secondly, that even though it might succeed well enough in this, one might still think that it had not succeeded in establishing or verifying its existence beyond any doubt. We need to consider other functions of language - other, that is, than mere propositions.

A basic assumption underlying the domesticity viewpoint is that there are indeed "real world" objects and events, enduring independently of our ideas. (For it might be thought that the failure of propositions to refer suggests there's nothing to refer to, in which case my argument would be pointless.) I do not for one moment doubt the existence of external objects, though the manner of their existence is controversial. So let's be quite clear, the question is, not whether (or how) those things exist, but whether the language we use can ever refer to them, in some way that ensures a "hard" connection with that realm.

One philosopher who was swayed by the conventional (Russellian) logic of the day was Gilbert Ryle. In his 1931 paper Systematically Misleading Expressions, he contended that some gramatically correct statements are not logically correct and need to be converted into a format that complies with classical logic (though his analysis dug much deeper than that). For example, because you can't refer to non-existant entities, he proposed that "Unicorns do not exist" be rendered as "Nothing is both a quadruped and herbivorous and the wearer of one horn (or whatever the marks of being an unicorn are)". (Strange that in the parenthesised phrase he appears to allow the unicorn some kind of existence after all!) It would never have occurred to Ryle that you can't successfully refer to entities that do exist, such as "Our village policeman" (his own example). Notice that the descriptive predicate in the paraphrased unicorn example doesn't mention unicorns at all: it is simply a substitution for "unicorn" in "Nothing is a unicorn", in which "unicorn" still looks like something that could be referred to. One might well ask whether the descriptive substitute predicate is of any help in abolishing the impression of referring.

So why do most philosophers still believe that many grammatically correct statements are not logically correct? Might it not be that the logical paradigm appealed to is incorrect? Why do reality, existence and reference take logical precedence over conceptualization - that is, in their publicly communicable form? If you make a model of a unicorn, what exactly is it a model of? What's the difference between saying "Tasmanian Tigers don't exist" and saying "Unicorns don't exist"? OK, there's a difference (Tasmanian Tigers once existed) but is it "systematically misleading" anybody? Perhaps more telling is our impartial use of the definite article as in "The Tasmanian Tiger no longer exists" and "The Unicorn never existed", underlining the idea that if "existing" and "real", "non-existent" and "not-real" are predicates (as I believe they are) we should be able to talk about entities that are not real as well as those that are (as discussed in Remarks on existence...).

Of course we perceive external objects and events from many different angles and in different contexts, therefore applying different descriptions, maybe even different names, to what we come to think of as being "the same" object. The domesticity argument implies that there's no single name or description which functions as a unique label, so to speak, "glued" to this object.

One might suppose (mistakenly) that every real object or person is a juxtaposition of concepts. Which ones we use to refer to anything is a matter of utility. Space, time, number etc are included amongst these concepts and do not stand aloof from them. Spatial form is obviously an important characteristic of most real things, but so are its spatial and temporal relations to other things and ourselves. Don’t we recognise these characteristics too? Yes, of course.

Yet although a person is clearly not just a bundle of concepts, whatever I know about a person is entirely conceptual. There's no point in trying to wholly identify a person by amassing the concepts I (or anybody) might be able to access in relation to the person.

To end this section, here are a few words that may help to mark the domestic/dictionary distinction:

domestic:  local, familiar, immediate, dynamic, hot, intense, rich, centred, focussed, elaborate, replete

dictionary:  broad, blanket, umbrella, cold, bare, sparse, bland, frigid, surface, covering

This list is intended to show that in general a restriction of context is accompanied by an enrichment of meaning, a principle which finds its most extreme realization in the modus operandi of some, but not all, proper names, as outlined in the following remarks.

Proper names and definite descriptions

The mountain of nonsense that has been written about proper names and definite descriptions beggars belief. Philosophers with logical and semantical leanings seem more interested in complying with artificial constructs than in fathoming how our minds work, in basic tasks like observation and recognition. Much of the trouble (with recent logical analysis) arises from the dubious idea that individuals must be treated quite differently from general features – as if recognising that Tom is Tom is somehow different from recognising that an albatross is an albatross. Just because proper names happen to apply to just one object (or collection) doesn’t make them especially different from any other descriptor. The primary service provided by any name is that one can recognise the thing(s) it stands for. It just happens that people are extremely complicated things, therefore recognition normally takes into account many factors. A certain smallish number often suffices, but the factors used by different people, or in different circumstances, may vary, and may not even overlap (see item 12 and ancillary notes). Occasional incorrect identification is inevitable. In principle, however, there’s no difference between recognising my wife or the Eiffel Tower and recognising the colour vermilion, the sea, a forest, a tree, a giraffe, a tower. And in logical arguments there's no good reason for handling proper names and universals* differently: both are elucidated by descriptions, often domestic in the one case, dictionary in the other, although, as we shall see, both names and descriptions can fall into either category or somewhere in between.

*Elsewhere I contend that universal propositions, as traditionally understood, are vacuous.

Why should we imagine the word mother to be any less "proper" than Mary Jones? Just because there’s a universal term mother that can be applied to most females, at some time in their lives? Nonsense! "Mother" (or an equivalent term) is most often used in specific contexts in each of which it has a specific meaning, which is different from its meaning in other contexts. Indeed there might be a common thread - one of the dictionary meanings of mother - running through the majority of usages of mother, but unless one is filling in forms (see below) this almost drops out of the picture. In practice a germinal domestic translation is most likely (see item 7).

Essentially, then, both descriptions and names, including proper names, only connote and do not actively denote, except perhaps in certain "on-locale" circumstances (see Anchored statements). Their behaviour in those circumstances, however, is no different from the behaviour of other elements of language, whether nouns, phrases containing verbs and adjectives or entire statements or commands. Otherwise the business of "standing for" some real object appears to be quite beyond their scope or capacity. A slightly different slant on this argument occurs in Existence: #6. This entry contains an example, re-printed below, which illustrates well the direction I'm coming from. (Adapted from Henri Bergson's Time and Free Will). Also see the supporting remarks in #3 of the same article.

A farmer has a flock of sheep. At first sight they might all look the same, but the farmer soon learns their differences and gives some of the sheep names, such as Matilda, Hop-along and Spotty. Some sheep, however, he cannot identify separately but only as classes, to which he gives names such as Woollens, Stubbies and Grandmas. The farmer also has some cattle. Is there any significant difference between his distiguishing the sheep from the cattle, the woollens from the stubbies, or Matilda from Hop-along? No. Is there any significant difference between calling certain farm animals sheep and cattle, woollens and stubbies, Matilda and Hop-along? No. (Do any of these names merit capital initial letters?) Is there any significant difference in the way proper names arise and the way universals arise? No.

Are there any terms or features of language that do denote? Not in the ordinary sense, of standing for objects that are otherwise connoted by those terms. It would seem that denoting (of a sort) can only be achieved externally to the use of descriptive language by making real links to objects, for example by pointing or demonstrating. This is presumably in part how we learnt the meanings of basic words, but, it goes without saying, the riddle of how children learn to speak and write is a labyrinth still fraught with questions, as is the riddle of how adults learn a second language, though perhaps to a lesser degree. There are only a few non-connotive words within a language that assist in making these external links, namely indexicals such as here and there, this and that, now and then. The sense of the indexical is, so to speak, extracted from the outside world. These words therefore do not normally denote in any ordinary sense; rather, they hunt for meaning. Nonetheless, as I shall point out later (#25), indexicals employed in this way are redundant.

As intimated above, proper names may be roughly classified as either domestic or dictionary or somewhere in between. Let's consider an example at the dictionary end of the scale, and I'll take Russell's much quoted "Venus" example. The domesticity approach holds that "the morning star" is one of many possible descriptions "referring" to a certain star, and "the evening star" is another description of the same star. But to what exactly are these descriptions "referring"? The answer can presumably only come in more words, such as a proper name or a dictionary definition. (A short definition of Venus might read "the sixth-largest planet in the solar system and the second in order from the sun".) The domesticity stance claims that the "reference" of the entity often described as "the morning star" or "the evening star" is just another description, but one which (in this instance) is more general than either of these; that is, it lies more towards the dictionary end of the scale. Or, if one says the "reference" is the name "Venus", then it is whatever one takes to be the meaning of "Venus", given that meanings are liguistic entities or perhaps mental impressions/models and cannot be identical with real objects. Having said that, in principle there's no particular reason why "the morning star" (or "the evening star") could not itself be a definition of Venus. (In contrast, according to the conventional approach, "the morning star is the same as the evening star" is an identity statement, the two definite descriptions picking out an object named Venus.)

The above illustration closely resembles the example of identifying my (fictitious) sister, Rebecca, by means of one or several descriptions, except that "Rebecca" could well be subject to a more domestic translation than various descriptions of her, depending very much on context.

So how do we deal with the question of the so-called "identity" of multiple descriptions applying to a single object or event, especially a named object. In fact there can be no identity as we have only meanings and concepts (not objects) at our disposal: "the morning star" does not mean the same as "the evening star" and neither means the same as "Venus" or any definition of Venus. Still, definitions there must be - any number of them - though really they are just bloated descriptions aimed at expanding the concept Venus to cover a wide range of situations. Let there be no misunderstanding: of course there's a real universe out there containing real stars and so on (see item 11). My objections are only (1) that the metaphysical x's and y's of the logical apparatus which supports the conventional approach to reference and identity fail dismally to answer our question, and (2) that the floating propositions of ordinary language also fall short of making a solid connection with real objects, events and situations. Except possibly in circumstances where the opportunity of using indexicals occurs (anchored statements), language cannot escape from itself.

If the above explanation seems insufficient to topple the conventional account of referring to an individual object in different situations or by means of different names, consider the following four scenarios:

A. The tree I'm looking at now is the same tree I'll be looking at in 10 seconds time.
Does this relate to one situation or two, one description of the tree or two? Indeed is it really the same tree I'm looking at and referring to all along? I'm inclined to say it's the same tree in two situations described by "now" and "in 10 seconds time". But that's because language, not the experience itself, is pulling the strings.

B. I'm looking at a tree now, but then I decide to take a short nap under the tree; after 15 minutes I wake up and see the tree again, albeit from a different angle.
Here the continuity of the experience is briefly interrupted, my personal aspect is different but otherwise the situation is much as in A. Indeed, there's nothing to choose between varying situation and varying aspect.

C. The bright star I see in the mornings is the same bright star I see in the evenings.
This is just an extension of B, with a longer blackout period and more appreciable change of aspect/situation. Furthermore the description has been generalised to cover multiple observations, and an injection of background knowledge about the planet Venus is presupposed. But evidently language is pulling the strings here just as it is in A (and no less coercively than the barrage from politicians, banks and Windows10 !). Real objects and experiences are the puppets of palaver.

D. The bright star I see in the morning is the same bright star that other people see in the morning.
Well, who am I talking to in A, B and C? Other people, hopefully, and not solely to myself (an impossibility, according to Wittgenstein). So if other people understand what I'm saying in those examples, they should understand just as well what I'm saying now: D is no more informative than C.

These remarks epitomise my belief that many of the problems of philosophy are fabrications ensuing from the way language organises our thoughts.

“The meaning of Rebecca is the person Rebecca”. I think Rebecca would go through life with an inferiority complex if she thought she was just a meaning. Something’s not quite right about using meaning this way. If asked for the “meaning” of Rebecca, I should not feel obliged to produce Rebecca, any more than I should feel obliged to produce Aristotle if asked for the meaning of Aristotle. But in Rebecca's case, it could be very helpful to do so.

“When Rebecca Robinson dies the person dies but not the meaning of Rebecca Robinson.” This is absurd too. It’s supposed to show that meaning and reference are not the same. Language never could take on board actual persons or objects, dead or alive, except maybe ostensively (i.e. by a direct observation, for example, as a replacement for a word or phrase, and even then the zzz is subjective). Though Aristotle be dead, we still "refer" to him. It's just that such reference consists only in whatever meaning we can muster to achieve our aims. But in this sense of "reference" one can refer both to dead people and to the characters in novels.

So if Rebecca herself is not the meaning of Rebecca, then what is? It’s natural to think of meanings as, or in terms of, ideas rather than entities, and this is certainly my perspective.

Anchored statements and labels

Broadly speaking, anchored statements can only be understood when uttered in the presence of the objects or events which they are about - usually either the subject of the statement as defined in Existence: #3 Aboutness) or a state of affairs or activity described by the whole statement. Typically the subject of an anchored statement is an indexical such as this or that, and this is what anchors it to an object or situation in the real world. It's arguable that in so doing the object effectively becomes included in the statement, but how it manages this feat is presumably for cognitive scientists to work out. It is not at all a simple matter. Suppose we invent some propositions in which a non-verbal object is directly included as subject:

X is a red cross
X is a "no entry" sign
X means "you are not allowed to go through any door marked with this sign"
X means "don't cross the road"
X is a big kiss for you
as opposed to:
Here's a big kiss for you: X (the "X" is external to the statement)
X marks the location of the treasure (X is part of the statement but not an introduced object)
The red cross means you're not allowed to go through the door (words replace the sign)

Clearly the possibilties of interpretation are many, and a prior understanding of grammatical features seems to be necessary. If an indexical is used to identify an extra-statemental object, as mentioned above an additional hurdle intervenes, namely the question of how the use of the indexical succeeds in singling out the object from its environment. (Probably the most common work performed by indexicals is to relate a statement or clause to another statement or clause, but this internal usage will not be considered here - see, I've just used two indexicals!)

A possible answer to this question is that indexicals don't single out objects from their environments, and, at least in that respect, they are redundant. For, if a statement situation is such that an indexical is capable of facilitating the identification of an object answering to some description, then the situation is such that no indexical is required. The pinpointing or tagging can be achieved with non-linguistic indicators such as gestures, noises, "semiotically" or simply by being on the spot. Examples are unlimited. Imagine watching a soccer match enlivened with a commentary, but, not being a fan of the game, you neither know the names of the players nor understand the jargon. "Giggs delivered a perfect cross to Beckham...." (sorry, MU fans, these notes are prehistoric!) Well, now you've got a better idea, even though the commentator did not (thank goodness) elaborate his spiel with "That's Ryan Giggs just making a pass called a cross to that player near the goal mouth and he's called David Beckham". Notice, here, that neither the proper names nor the statement as a whole denote any person or activity; no object or event is identified by virtue of the sense of the words employed. On the contrary, this is effectively a naming situation, in which meaning is transferred from the situation to the words.

Be that as it may, none of the above deliberations explains how statements direct our attention to particular features in our environment. It seems to depend on our expectations and knowing what sort of language applies to what:- names of a certain kind apply to people, verbs often apply to actions and so on. In passing a jewellery cabinet in which there is displayed a green emerald the tour director says "This is a green emerald" and adds "Most emeralds are indeed green". If a little tipsy she might have said instead "This meralda is glupper, most meraldas are in-in-indeed glupper", and because the grammar is correct those in the party with no knowledge of jewellery, or little knowledge of English, might have accepted her announcement, only to discover later that there's no such thing as a glupper meralda. If more than just a little tipsy she might have said "Unkitunk woozlepop" which doubtless would have left everybody totally confused. Or she might have pointed to the emerald and chimed "Dong!" which, translated into English, probably had an intended meaning something like "Look there!"

Well, this is just a crazy way of exemplifying the obvious: anchored statements and activities do not function in isolation, but draw upon various implicit background factors, circumstances/context, repeated use, analogous usages/instances, intensions of the agent and expectations of the heeder and so on. [to be continued]

26. Labels.
Labels (of the sort I'm thinking of) are very like anchored statements:- a label with "Rebecca Robinson" written on it, stuck onto a person who is indeed called Rebecca Robinson, says the same as the statement "This is Rebecca Robinson" made in the presence of that person (see below). And like those statements it can be either a naming label or a referring label. The difference is only that "naming" is an initial "referring", at least for someone, while referring as such (using a name) can occur at any time. Labels generally work better than anchors, but are not perfect - sometimes it depends where you stick the label! A sign at the edge of the sea saying "Pacific Ocean" would not be very helpful.

27. Naming.
Surely an initial naming, at least, must be a naming of something that can be described, by whatever means. A name must identify a person, thing or event, which ultimately depends on observation, hence, one might suppose, description. In general, one should be able to recognise the thing named, somehow or other - well, this is often impossible (see #28). Still, that does not imply that a name must be synonymous with a description.

It takes little insight to appreciate that the conferment or emergence of a name is not an especially remarkable or unique kind of process. Let's run the argument backwards, so to speak, by considering general nouns first (and there are no grounds for not including other parts of speech as well). Clearly, most general names develop over a very long time, almost as a language itself develops. They have no distinct point of inception. But some general names do arise quite abruptly or over a short period - as when a new species of plant or animal is discovered, new technical terms are invented or new names are applied to old ideas. In principle there's no case for assigning singular names (for example the name of a mountain, river or island) to a different logical category from general nouns - firstly because recognition, not number, is the key to understanding the naming process, and secondly because general nouns must be answerable to universal propositions, which, as explained elsewhere (e.g. here and here), belong in the realm of mythical creatures. As for the "given names" of people, well, the only reason why they appear suddenly is because people are born suddenly.

The privileged status of naming also faces the charge of arbitrariness, particularly with respect to the circumstances where a first naming for one person may not be for another. There's little difference between naming (for the first time) and hearing a name for the first time. A baptismal naming might be a cerebral invention, but once revealed it is just another referring phrase, normally anchored, a first-time usage uttered in the presence of the person named, though there's no special reason for excluding floating usages from the picture. Suppose, for example, Mary's husband phoned his mother with the exciting news "Mum, you've just got a new grand daughter called Elspeth Veronica", and that was the very first announcement of the name. Again, subsequent disclosures of the name differ hardly at all from the initial announcement. In either situation, there's normally a person who knows the name and utters it and a person who doesn't know the name on the receiving end. The soccer match example given in #24 typifies this scenario.

Finally, the inherent arbitrariness of naming becomes very evident when it occurs in informal contexts. The way that nicknames and similar appellations arise (and fade away) is capricious and undefined, and often limited to particular social circles. This aspect of naming is exemplified in the Ancillary remarks below.

28. Formal identification.
For many purposes, the production of certain authenticated documents is sufficient to formally prove identity. Official documents, usually witnessed and bearing signatures, such as birth certificates, passports, driving licences, citizenship certificates, social security cards, marriage certificates, education certificates, medical cards, bank cards, death certificates, even town rates notices and electricity bills - all of these provide evidence of personal continuity and significant events throughout life, including beginning and end. First let us note that documents can be faked or stolen, and in that respect have the same weaknesses as ordinary material propositions (See #31 - and the real world is full of fakes too.) And secondly, while it might be held that many such documents relate to typical naming situations (as in #26), we should remember that names can be applied to fictional entities and objects which occur only as hypotheses, representations and the like, and which may never come into existence at all. Things of this kind may also be well documented (for example, the blueprints for a nuclear submarine that is never built), as may plans for events (such as an abandoned cricket match).

However, it's arguable that genuine identity records must ultimately refer to "naming" events, or something similar, and certain philosophers appear to model their theories of reference on documents of this type. That is, they base their concept of referring on methods of formal identification.

I might add that formal ID documents are not labels, attached to our bodies by straps or glue. We may carry them around with us but still they don't label us and no alarm bells ring when they leave our hands.

Well, I may not have a sister named Rebecca, but I do have a wife, two sons and three brothers. Whether or not they will appreciate my optimistic attempt to inject new life into a hugely overworked cliché "You mean much more to me than words can express", I trust they will happily accept and cradle those words along with the deep affection they convey.

"I'm putting some pepper in the soup" is true if I'm putting some pepper in the soup, and that's that - so they say. In this sentence, however, you cannot understand the unquoted (antecedent) clause as being a clause, a sentence, a proposition or anything else. It ought not to be said at all, it simply is not another level of language, as Tarski might have suggested. This, I think, is an important part of the answer to the "connection" problem. Language may be the only way to express something, at least in print. But as I see it, the unquoted clause is redundant, to be replaced by an activity: "I'm putting some pepper in the soup" is true if ...[present action of me putting pepper in soup]. We seem to lack the means of talking about something without talking! The "something" is normally either a non-verbal experience or a real-world object/activity.

Yet in an actual "Putting pepper in the soup" situation a person with normal eyesight who understands English would hardly need even the first (quoted) statement, let alone the unquoted version. It would be odd if we ever had occasion to say: " 'I'm putting some pepper in the soup' is true because that's what I'm doing. Can't you see what I'm doing, you idiot?" (In this instance only a child might want to know as part of a learning experience - see #31.) The procedure could be followed perfectly well without needing any narration at all, so there would be no statement to be true or false. Nonetheless the statement "I'm putting some pepper in the soup" would be true if I uttered it. So would a proposition such as "At [such and such a time and place] Dave Robinson was putting pepper in some soup".

TV chefs, of whom there's an indigestible surfeit, always tell us what they're doing, partly because they love to blather on but mainly because their viewers can't always observe their procedures easily and/or don't know the names of the ingredients they're using. Whether inspired by culinary capers or not, this kind of engagement, as we saw in #24, appears to create a firm but fleeting connection between propositions and real-world objects and events, even without the aid of indexicals. "Fleeting", I add, because subsequent links will invariably be internal to verbal communication.

Turning to on-locale requests and commands, the connection with reality often seems to be stronger still, because they demand interaction with particular aspects of one's surroundings. Let's begin with a Tai Chi class (because that's what happens in the park opposite our house every Tuesday and Saturday). Normally the participants just copy the actions of the instructor, but occasionally she might beg "Do this! (note the indexical) or "Raise your arms a little higher" (note the lack of an indexical). All three methods are equally efficient, suggesting once again that sometimes language is redundant. As we saw in #25, the possibility of redundancy arises just in those circumstances where anchored statements (or commands) might be employed. But in most situations language is either essential (as in the soccer match example above) or can greatly assist in conveying one's intentions, as in these commands: "Pass the spaghetti, please" (a polite request driven by hunger), "Shoot the bastard" (an angry outburst very likely with dire consequences), "Meet me in the park in 10 minutes" (a partly ex situ overture portending further adventure), "Fetch the ball!" (an order even dogs can understand). All are better handled with words than signals and in a way they all resemble ordinary domestic statements, but the degree of "domesticity" depends on the situation rather than the content of the command.

Finally we come up up against the disease that bugs most of the above deliberations. Any hope of ensuring the anchoring of a statement to an object is dispelled by the elementary observation that, for one reason or another, the statement (or expression) may be misleading or the object may be fake, or one's perception of it mistaken! These impediments are commonplace. As Nine TV network host Lisa Wilkinson puts it (referring to social media, 30/06/2017), "There's so much white noise out there", adding that journalism is important because it has the capacity to find "real facts, not fake news". Hmm, shades of Donald Trump here! Indeed the consequences of deceptive or misinterpreted signals can lead to disaster. These considerations alone suggest that there are no cast-iron links between language and the real world. Reference and material truths have no safe harbour.

Ancillary remarks on proper names and definite descriptions

Some time ago we bought two chickens which we named Seema and Neema. After a year or so we moved house and sold the chickens. The new owners decided they would call them Curry and Rice. The only way we could tell which was which was by their features. So the names Seema and Neema served only as tools of recognition. “Go and catch Seema” meant something like “Go and catch the plain brown chicken with the larger comb”. “Go and catch Neema” meant something like “Go and catch the speckledy brown chicken with the smaller comb”. That was in the context of our family and back yard. When the new owners used the names Curry and Rice they were referring to the same chickens in the context of their family and back yard.

But now the question whether the names stand for “descriptions”*, i.e. a sentence or two about the identifying features of each chicken, or whether they stand for some kind of “meaningless” labels rigidly attached to the chickens – this question is absurd. The first option seems better, for clearly we could, if we needed to, verbally point out the best way of distinguishing between the two chickens. Yet clearly the names don’t stand for those descriptions. Recognising Seema and Neema does not depend on verbalising their features. (I might have ventured to suggest the names stand for the chickens, but I'm certainly not going down that track!). But nor are they “empty” tags with no “meaningful” connections. For we know that “Catch Seema” tells us to go and look for the plain brown chicken with the larger comb. And the tags are not permanent – the new owners used different tags.

In fact the names are used just like any other noun – or adjective for that matter. When we consider the word “chicken” or “oats” or “clucking” we don’t get into arguments about whether they stand for descriptions or objects. Nor do we worry should taxonomists, or common usage, invent new names for species or gadgets etc. (or the fact that in countries with different languages different words will invariably be used). We don’t insist that “red” or “grass” must be equivalent to some (presumably more expansive) description. What would be the point? You could go on elaborating forever, or until your knowledge runs out. The utility of red, grass, chickens, my chickens, Seema, The Eiffel Tower lies in the mental impressions we get and the responses we make when hearing the words in various contexts. But the origins of the words may vary immensely, and mostly would have followed a long course with many influences.

The use of names for the chickens is of course very clannish – domestic is the term I've been using. Maybe only our family understands them. However, there are basically two ways of widening the circle. I could explain to other people who Seema and Neema are descriptively, or I could take them into the back yard and point them out; in that case I might not need to say anything at all, for, if they are observant, they could just see which is which. True, they might hook their observations onto some verbal thoughts as a memory aid, but for most people I suspect this would be unnecessary (indeed memory trainers often tell you to do just the opposite – hook your verbal thoughts onto images).

Now lets go one step back again. Suppose I am a hermit living alone with two chickens (how often I wish it were so!). I call my chickens “Seema” and “Neema” but nobody else knows what I call them. I use these names just for pleasure, or to make it easier to talk to myself: “Neema hasn’t laid an egg for awhile so maybe she’s got worms ...” Although some philosophers (whom I regard as a little weird) claim that private languages don’t exist, I’d say I’m fully justified in maintaining that “Seema” and “Neema” are here elements in my private language, or (if that’s too much for some academics to swallow) at least that “Seema” and “Neema” are private elements in my language. The names may never be spoken or recorded, but they remain useful to me. (However, they are “recorded” as brain traces.)

If you are “clued up”, you’ll probably take one of two paths from here. Either you’ll immediately reject my thought-words “Seema” and “Neema” as not being proper names (for whatever reason – some philosophers might argue that the script is non-existent until it’s out of my head) or you’ll accept them as proper names and perhaps linger a little longer with them. But where next?

We could step back yet again, one way or another. Suppose Seema and Neema don’t exist at all: they are just part of my fantasy world, a world I might have owned as a child. Or suppose the script does exist, but it is only a novel. (Well, I’d rather be wondering how the names function here than in some trumped up twin earth or possible world!) In either case nothing real corresponds to the names, so one might say they are not “grounded”. They exist only in thought or in print. They could be amplified by description but not ostensively e.g. by pointing or confrontation. Obviously such names would fall through the rigid designator net, and for that matter any net that requires names to denote or refer to real objects.

Really, it doesn’t matter much what theory of proper names one latches onto. You can dream up whatever meaning of “meaning” you like, bend it into any shape that suits your purpose, but it would be helpful to settle for a plausible meaning of “meaning” - one that has a friendly ring to it. It does matter that far-fetched arguments are brought to their knees.

(See the "red gremlin" example in A ravenous tale. If there are more than one roaming around they can only be distinguished by description, unless captured or encountered face-to-face.)

More on the meaning of names

Of names in general, a kind of “picture” theory is the “common-sense” choice. Often a name used in a broader sense takes in a family of objects that resemble each other, based on appearance or function. For example, the name “tile” might include roof tiles, wall tiles, floor tiles, capping tiles.... An even wider usage might cover things like scrabble and mah-jong tiles, but of course any particular context is unlikely to include all these in its scope. In a narrower context only one kind of tile might be meant. At each level, though, it’s clear that the word “tile” is public and communicative. How a person understands the word depends on what he expects tiles to look like in the current context. His mind may at first be open to a wide range of possibilities, but will quickly narrow these down when he understands the context. The “meaning” of tile seems to refer primarily to its public usage and secondarily to one’s personal “picture” or expectancy. How that picture develops is a psychological rather than a philosophical problem. But normally a person’s “picture” – what he believes the word means – is what he believes to be its public usage (in various contexts or in a particular context). If he did not believe that, the word would have no use for him in the public world. So there seems to be a fair case for divesting “meaning” of its subjective cloak and planting it in the public arena. (See note on language games below).

Well, it all depends:- those who want to externalise meaning would doubtless say the “meaning” of a Beethoven symphony is the score; those who think meaning is discerned in people’s behaviour would say it consists in the scraping, blowing and thumping activities of the orchestra. Certain realist-empiricists might believe it relates to the sound waves produced by the orchestra’s scraping, blowing and thumping. And somewhere in here we have to fit CDs and other recording media. Isn’t it obvious that none of these camps can be right? The “meaning” of the symphony is more closely related to what we hear and what Beethoven intended us to hear. With music, however, one cannot easily distinguish what we hear from the “meaning” of what we hear. The next, and last, step up is how we react - how it affects our emotions. And this is also true of verbal language, though the possibilities of reaction are much wider.

However, if the meaning of a name is considered to be its reference, and the reference is an external entity, then meaning has been externalised. (And according to the causative theory of the meaning of proper names, the meaning goes back to causally connected roots.) But if this view is extended to natural kinds, I would presume that the meaning of gold should be particular lumps of gold, and not its scientific properties such as atomic number, stability, melting point, conductivity... And if there are causal connections involved I would think historical roots would be more relevant than scientific (analytic) ones.

.....More oddments:

It’s arguable that both real existence and singularity can be included in the meaning of a name.

While many different singular objects (especially persons) may have the same name, this wholly uninteresting state of affairs is indistinguishable from that in which any word has various meanings (related or not). It's hard to believe anyone would suggest names such as "Smith", "Brown".... are predicates on the grounds that their owners belong to the class of Smiths, the class of Browns.... But I could name at least one philosopher who has published a paper to this effect.

Note Wittgenstein’s language games, a concept which more nearly succeeds in bridging the gap between language and actuality. But in many respects rather obvious and naive, e.g. words like “game” with related usage but unrelated extremes, are essentially just another case of words having various meanings. No big deal.

How the word "sheep" acquired its meaning interrogates the distant past. But I can't believe there's any significant difference between how "sheep" acquired its meaning and how "France" or "Aristotle" acquired their meanings. Just that the process took longer. And perhaps I should add “gold” and “water” to these examples, as natural kinds seem to pose special problems for some academics. Why I don’t know. (It has lead to some doubtful suggestions, such as that meanings are objective.)

Colour examples: "This raven is black", "This emerald is green" > "(Most) ravens are black", "(Most) emeralds are green". The first question is: in pointing to a raven and uttering "This raven is black", why didn't I say "This gibble is blupper", or worse, "Unkitunk woozlepop" or just "Ug!" which, translated into English, means something like "Look there!"? Well, I suppose if that’s what I said then, that’s what you might well be saying now, when referring to one of those big black croaking flying things!

If you’re not a birder you might be perpetually mis-identifying species. After a bird spotter has identified a bird for you as being a bar-tailed godwit, you might believe that any wader was properly called by that name. You’d only be partly wrong. If it’s good enough for you, it’s good enough for them.

If you join a birding group and somebody points out and names a species of bird you’ve never seen before, isn’t that a “baptismal naming” as far as you’re concerned?

I'm not sure when and why I trumped up the following, or whether it has any bearing on the present narrative.

In everyday language there are only three kinds of general proposition, in simple terms:

(1) Contingent
If you have found or will find an A in a closed collection Q, it had or will have the attribute a. (2) Predictive
If at some future time you find an A in circumstances Q, A is likely to have the attribute a, based on past observations of A-like things in Q-like circumstances.

(3) Definitional
If you find a convergence, x, of attributes a, b, c... in a context of type Q you’re obliged to call x an A
in contexts Q.
          It makes very little difference if there is only one x.
          It makes very little difference if “A” is an initial naming of x or a renaming of x.
          Using A for x in Q does not preclude the use of another name for x in Q.
          In a context other than Q, A may name some convergence, y, different from x.
          “Attributes” includes external relations, both spatial and temporal.
          A, even if a “proper name”, may be replaced by a description representing a sufficient convergence
          of attributes a,b,c...
Here I would argue that the only way of identifying anything is by its attributes (including external relations). Cases of mistaken identity don’t alter the truth of this.
Note: identification in this way may not confirm the existence of A, but only accommodate A, in my sense of accommodation . Thus all the known circumstances of A’s baptism could be a myth.
(“That’s not my mother’s driving licence, it’s Mrs Johnson’s” is not very different from “Shakespeare didn’t write “Shakespeare’s” plays, Marlowe wrote them”).

What have we learnt, if anything? Some words got their meanings a very long time ago and might have acquired them slowly. Others got their meaning recently and quickly. Some get it almost instantly by naming. But in general, from a philosophical point of view, how words acquire meaning is about as interesting as how Matilda acquired her hat. So much for this article!