Click here and press the right key for the next slide (or swipe left)

also ...

Press the left key to go backwards (or swipe right)

Press n to toggle whether notes are shown (or add '?notes' to the url before the #)

Press m or double tap to slide thumbnails (menu)

Press ? at any time to show the keyboard shortcuts

\title {Words and Things \\ Lecture 06}

\maketitle

Lecture 06:

Words and Things

[email protected]

\def \ititle {Lecture 06}

\def \isubtitle {Words and Things}

\begin{center}

{\Large

\textbf{\ititle}: \isubtitle

}

\iemail %

\end{center}

\section{Referential vs Attributive uses}

Recall Russell’s Theory of Descriptions ...

The question Russell’s theory was supposed to answer was,

Which words (and phrases) refer?

To understand this question, it is helpful to distinguish communicating about something from referring to it. You can communicate about something without referring to it.

Distinction: communicating about vs referring

For example, suppose Ayesha is eating lunch really loudly. Then you can communicate about Ayesha in two ways. You might just utter ‘Ayesha is eating lunch’, or you might be more coy and say ‘*someone* [emphasis] is eating lunch’ while raising your eyebrows.

In this second case you are communicating about Ayesha without referring to her. Strictly speaking, your utterance would be true even if it were not Ayesha but someone else who was eating lunch.

[They eat lunch in this lecture so you can use the example to tease a noisy eater.]

‘Ayesha is eating lunch’

‘Someone is eating lunch’

So your utterance of ‘Ayesha’ refers to Ayesha ...

... whereas your utterance of ‘someone’ does not refer at all.

Russell’s theory is about the contrast between another pair of examples ... Given that Ayesha is the smelliest cat in my house, you might think that there’s really no interesting distinction between these two utterances.

‘Ayesha is fluffy’

‘The smelliest cat in my house is fluffy’

But whereas most would think of the utterance of ‘Ayesha’ as referring to Ayesha,

Russell thought that an utterance of ‘the smelliest cat in my house’ does not refer, in just the way that an utterance of ‘someone’ does not refer.

Let’s take a look at Russell’s theory

‘we may put, in place of “the smelliest cat in my house,” the following: “One and only one entity is the smelliest cat in my house, and that cat is fluffy”’

Russell, 1905 p. 488

There is a tiny complication here ...

As you might remember, Russell also thought that proper names like ‘Ayesha’ are really disguised descriptions.

Three applications

Analysis of definite descriptions

Analysis of proper names (they are usually really disguised descriptions)

Distinguishing knowledge by description from knowledge by acquaintance

Here we are ignoring that part of his view, which I want us to treat as a separate debate. For now we are focussing on Russell’s attempt to characterise phrases like ‘the smelliest cat in my house’.

When we discussed Russell’s theory of descriptions, we saw that there were some good objections to it. However, we also saw that Ludlow has isolated a core idea which seems to withstand the objections.

‘The critical question is whether the sentences in which [phrases of the form ‘the so-and-so’] appear are quantificational or referential, and Russell may well be right about the critical cases here.

That is, many apparently referential constructions may in fact be quantificational.’

Ludlow, 2004

The core insight (‘many apparently referential constructions may in fact be quantificational’) is exactly what we are about to challenge in Donnellan ...

Donnellan’s objection

Today I want to look at another objection to Russell’s theory, one due to Donnellan. It is not a good objection, but it continues to be very influential. And, as you’ll see if you are very patient, reflecting on the objection can take us somewhere interesting even if, as I think, the objection fails.

“Suppose one is at a party and, seeing an interesting-looking person holding a Martini glass, one asks, “Who is the man drinking a Martini?” If it should turn out that there is only water in the glass, one has nevertheless asked a question about a particular person, a question that it is possible for someone to answer”

\citep[p.~287]{Donnellan:1966th}.

(Donnellan 1966: 287).

Ask them, What do you think Donnellan’s objection might be? Can you anticipate an objection based on this assertion? Turn to the person next to you. You have 90 seconds to discuss this.

[After the discussion in pairs, ask them to make suggestions in a whole-class discussion.]

The argument from misdescription (Donnellan’s?)

If ‘the man drinking a Martini’ were quantificational, then the question would not be about anyone at all.

In that case it would be puzzling how the question can be answered.

But the question is easy to answer.

So the ‘the man drinking a Martini’ can’t be quantificational.

It must be referential.

The argument from misdescription (Donnellan’s?) \begin{enumerate} \item If ‘the man drinking a Martini’ were quantificational, then the question would not be about anyone at all. \item In that case it would be puzzling how the question can be answered. \item But the question is easy to answer. \item So the ‘the man drinking a Martini’ can’t be quantificational. \item Therefore it must be referential. \end{enumerate}

The question Russell’s theory was supposed to answer was,

Which words (and phrases) refer?

Russell: Utterances of phrases of the form ‘the man drinking Martini’ are always quantificational.

Donnellan: Utterances of phrases of the form ‘the man drinking Martini’ are sometimes referential.

Note that Donnellan doesn’t think all utterances of phrases of the form ‘the man drinking Martini’ are referential. Instead he thinks ...

That there some are referential and some are quantificational.

Distinction: utterances of ‘the man drinking Martini’ can be ...

referential

[‘Who is the man drinking Martini?’]

attributive

[‘I have no idea who it is, but one man here is drinking Martini. Whoever he is, I want the man drinking Martini thrown out of this party.’]

“I will call the two uses of definite descriptions I have in mind the attributive use and the referential use.

A speaker who uses a definite description attributively in an assertion states something about whoever or whatever is the so-and-so.

A speaker who uses a definite description referentially in an assertion, on the other hand, uses the description to enable his audience to pick out whom or what he is talking about and states something about that person or thing”

\citep[p.~285]{Donnellan:1966th}.

Donnellan, 1966 p. 285

Note that Donnellan is talking about uses, not about words or phrases. Throughout this course I have been stressing a distinction between the idea that utterances of words refer and the idea that words themselves refer.

How does Donnellan’s talk about ‘uses’ fit with this distinction? I think what he might as well be talking about are utterances. In my terms, his idea is that some utterances of definite descriptions refer whereas others do not. The former are the ‘referential’ utterances, the latter are the ‘attributive’ utterances.

Consider a second example which might (has) been used to make Donellan’s distinction (see \citet{ludlow:1991_indefinite} for details) ...

‘The man drinking Martini is wearing a hat.’

‘No he isn’t, and he is not drinking Martini.’

What do the utterances of ‘he’ refer to? It seems they refer to the very man the utterance of the first sentence (‘The man drinking Martini is wearing a hat’) was about. And this could be taken to indicate that the utterance of ‘the man drinking Martini’ also refers to this man. At least I suppose Donnellan would think this. (This is far from obvious; \citet{ludlow:1991_indefinite} mention a range of possibilities).

The question Russell’s theory was supposed to answer was,

Which words (and phrases) refer?

Russell: Utterances of phrases of the form ‘the man drinking Martini’ are always quantificational.

Donnellan: Utterances of phrases of the form ‘the man drinking Martini’ are sometimes referential.

Who is right, Russell or Donnellan?

[after they’ve discussed in pairs, then in whole group] In order to further evaluate his argument, we need to consider how his observations might fit into a broader pattern of observations of how communication by language works ...

Pragmatic Aspects

\section{Pragmatic Aspects}

Consider a fairly standard picture of how communication with words might work:

\emph{A Simple Picture:} \begin{enumerate} \item 1. There is a sentence \item 2. which you utter in a context, and \item 3. the sentence plus the context fixes what the words refer to in accordance with codifiable rules. \end{enumerate} Example, when you utter ‘I have had breakfast’, your utterance of ‘I’ refers to you because you are the utterer.

A Simple Picture

1. There is a sentence, e.g. ‘I have had breakfast’,

2. which you utter in a context, and

3. the sentence plus the context fixes what the words refer to in accordance with codifiable rules.
E.g. ‘I’ refers to the utterer.

But things are not are quite this simple. Take an utterance of ‘I have had breakfast’. Suppose you say this in response to my question, ‘Are you hungry’. As \citet{Carston:2002tl} notes, I would normally understand that you’ve had breakfast \emph{today}.

‘I have had the oil changed’

Now contrast other sentences which have the same form. An utterance of ‘I have had the oil changed’ would not normally convey anything about today.

‘I have had a kidney removed’

And some utterances of the same form convey once-in-a-lifetime events.

‘I have had fermented fish for breakfast’

Finally, some utterance might convey quite different things depending on where in the world you are. As a Norwegian in Norway, you might reasonably be interpreted as talking about today; as a North American recounting a visit, what you communicate would probably not be about today.

[on leaving a party] ‘I have had great evening. This wasn’t it.’ (Groucho Marx)

Finally, as Grice notes, what your audience understands can be modified by things you go on to say. For example ...

What can we conclude?

If we accept the Simple Picture,

then what utterances communicate goes beyond

what any words and phrases refer to.

This is far from an isolated example ...

Jack and Jill went up the hill [together].

Beatrice has always been a great lecturer [since she’s been a lecturer, not since she was born].

Sigrid took out her key and opened the door [with her key].

Harold went to the edge of the pool and jumped [into the pool].

Carston, 2002 p. 22

What you have just been looking at are all examples of what is called ‘free enrichment’

“In many instances, it seems that the pragmatic contribution to the proposition expressed by an utterance goes well beyond ensuring minimal propositionality.”

\citep[p.~22]{Carston:2002tl}

Carston, 2002 p. 22

This is a pervasive feature of communication by language.

What can we conclude? As I said before,

If we accept the Simple Picture,

then what utterances communicate goes beyond

what any words and phrases refer to.

\citet[p.~75]{Neale:1990hg} proposes a three-fold distinction (to which \citet{ludlow:1991_indefinite} add SG): \begin{enumerate} \item MS, the meaning of the sentence; \item PE, the proposition expressed; \item PM, the proposition meant; \item SG, the speaker’s grounds for making an utterance. \end{enumerate}

MS, the meaning of the sentence;

PE, the proposition expressed; and

PM, the proposition meant.

Neale, 1990 p. 75

The meaning of the sentence is supposed to be something the sentence has irrespective of any particular context of utterance. This is not something that enables us to evaluate whether the sentence is true or false since the meaning of the sentence does not determine who the sentence is about. Accordingly no proposition is involved at all.

‘I have had breakfast’

To illustrate, consider the sentence ‘I have had breakfast’ again. Is this sentence true or false? Without any context of utterance, there is no saying who it is about. So the sentence cannot be either true or false. So MS cannot be a proposition, nor anything capable of being true or false.

We get to PE, the proposition expressed, by adding in the context of utterance. In this case, the proposition expressed is ‘I [Ayesha] have had breakfast.’ And this proposition is true just if Ayesha’s life has not been entirely breakfastless.

When people talk about ‘literal meaning’, they sometimes have in mind the PE.

The Proposition Meant, PM, is what Ayesha intends to convey in uttering this sentence.

Typically for every PE there will be many PM. Depending on the context these may include: I [Ayesha] have had breakfast today [Tuesday]. I [Ayesha] am not hungry. I [Ayesha] do not wish to have breakfast with you [Steve/John] now [Thursday 13.55pm].

Is the distinction clear to you? Do you have any questions about it? Do you think the distinction is correct?

[After discussion] Note that this is merely a schematic proposal. We are leaving open some important questions. For example, How can we rigorously distinguish PM from PE? And, How is PM determined?

If we wanted to rely heavily on this three-fold distinction, we would need to answer these questions, and more.

Come to think of it, why should we accept that sentences have anything like meanings at all? I’ve been stressing that we should anchor our discussion of technical terms like ‘reference’ and ‘meaning’ in facts to be explained.

Take a moment to recall the original justification for introducing the idea that utterances of words and phrases can refer to things.

Ayesha utters ‘Earth is being warmed by human activity.’

Beatrice utters ‘Mars is being warmed by human activity.’

ftbe: Those utterances differ in that one is made true by how things are with Earth whereas the other is made true by how things are with Mars.

What (if anything) is reference?

If this justifies anything, it justifies thinking that some utterances of words and phrases refer.

What could justify holding that
words, phrases or sentences
have meanings?

[I’m not going to discuss this question further in this lecture: the point of the discussion is just to get them thinking about the question which we’ll come back to in the next lecture.]

MS, the meaning of the sentence;

PE, the proposition expressed; and

PM, the proposition meant.

Neale, 1990 p. 75

So what is my point? I’m going to take a leap. Even without any proper understanding of why we should think of sentences as having meanings or of there being a distinction between PE and PM, I’m going to assume, just for now, that Neale’s three-fold distinction is correct. And we’re going to continue as if we all understood what these \emph{meanings} and \emph{propositions expressed} are (although I’m a little doubtful of this).

In terms of this distinction, Carston’s observations about free enrichment show that there is a significant gap between PE and PM. Consider one last contrast as an example ...

‘There is milk in the fridge.’

As \citet[p.~24]{Carston:2003os} notes, the PE is a very weak general proposition: that there is some presence of milk in the fridge (perhaps just a stale drip or two on a shelf). By contrast, the PM is much more specific: for instance, that there is milk usable for coffee in the fridge. Arguably, it is only the PM that is communicated; the PE is something competent communicators might hardly be aware of. ... and, arguably, it is only these latter that are meant”

‘There is mould in the fridge.’

Contrast the milk utterance with an utterance of ‘There is mould in the fridge’. You can see that the utterance of a sentence of the same form might involve a different PM: what I’m communicating this time is (hopefully) just about the merest traces of mould. (But actually maybe not. I suppose it depends on your standards.)

‘There are nuts in the fridge.’

Finally, an utterance of the same sentence might be linked to different PM in different contexts. For instance, if I was hungry, your utterance of ‘There are nuts in the fridge’ would suggest more than a trace elements; but if we were in a food factory where concern about allergies is an issue, an utterance of the same sentence might convey only that there are at least traces of nut in the fridge.

Recall the fairly standard picture of how communication with words might work.

The discussion so far does not show that the Simple Picture is incorrect, but it does show that it is importantly incomplete.

Why? Because the Simple Picture only captures PE, not PM. And if we want to understand communication with words, then of course it is PM that ultimately matters. So we can’t rely on a picture in which PM does not appear at all.

Now let’s go back to Donnellan’s objection to Russell on descriptions and consider how these ideas might apply ...

A Simple Picture

1. There is a sentence, e.g. ‘There is milk in the fridge’,

2. which you utter in a context, and

3. the sentence plus the context fixes what the words refer to in accordance with codifiable rules.

Pragmatics vs Donnellan

\section{Pragmatics vs Donnellan}

The question Russell’s theory was supposed to answer was,

Which words (and phrases) refer?

Russell: Utterances of phrases of the form ‘the man drinking Martini’ are always quantificational.

Donnellan: Utterances of phrases of the form ‘the man drinking Martini’ are sometimes referential.

Recall Donnellan’s disagreement with Russell ...

What was the argument for Donnellan’s position? It was based on examples like this one ...

‘The man drinking Martini is wearing a hat.’

‘No he isn’t, and he is not drinking Martini.’

Donnellan’s idea was that this. The conversation is successful although, as it turns out, the man is not drinking Martini. Donnellan takes this to show that the utterance of the phrase ‘The man drinking Martini’ is referential rather than quantificational.

An alternative to Donnellan’s view:

The PE is ‘There is one and only one man drinking Martini, and he is wearing a hat’.

The PM is a proposition about that man (who isn’t actually drinking Martini).

An alternative to Donnellan’s view: \begin{enumerate} \item The PE is ‘There is one and only one man drinking Martini, and he is wearing a hat’. \item The PM is a proposition about that man (who isn’t actually drinking Martini). \end{enumerate} Given this alternative, can Donnellan’s objection to Russell on descriptions be rejected on the grounds that his distinction between attributive and referential uses concerns the Proposition Meant (PM) whereas Russell’s theory of descriptions is about the Proposition Expressed (PE)?

That might be too quick---it depends on how anaphoric uses of pronouns work, among other things (see \citet{ludlow:1991_indefinite}).

But that is a topic beyond the scope of these lectures (although those interested can read further, and it would be a fine essay topic).

Neale provides a nice clear statement of the disagreement as it now stands:

“The Russellian and the ambiguity theorist [i.e. Donnellan] agree that when a description is used referentially, (one of) the proposition(s) meant is object-dependent;

they just provide different explanations of this fact.

The referentialist complicates the semantics of ‘the’; [i.e. explains it by appeal to PE]

the Russellian appeals to antecedently motivated principles governing the nature of rational discourse and ordinary inference [i.e. explains it by appeal to PM].”

\citep[p.~90]{Neale:1990hg}

Neale, 1990 p. 90

conclusion

In conclusion, ...

So what have we learned? Three things.

There may a distinction between MS, PE and PM.

In evaulating objections to Russell’s theory of descriptions, we should to take this three-fold distinction into account.

... This is a bit of an understatement. In evaulating almost any theory about what words or phrases mean, we will probably need that distinction. (This is going to be a topic in future lectures.)

To better understand communication with words, we probably need to look more deeply at the distinction between MS, PE and PM.

And that could be the end of this lecture. There’s just one last thing I’d like to squeeze in if there’s time ...

The Meaning of a Sentence, the Proposition Expressed and the Proposition Meant

\section{The Meaning of a Sentence, the Proposition Expressed and the Proposition Meant}

Recall this distinction which we provisionally accepted and used in considering Donnellan ...

MS, the meaning of the sentence;

PE, the proposition expressed; and

PM, the proposition meant.

Neale, 1990 p. 75

Why accept that sentences have meanings? And what are meanings anyway?

What are the meanings of sentences?

Why suppose that sentences have meanings?

Two questions about MS: \begin{enumerate} \item What are the meanings of sentences? \item Why suppose that sentences have meanings? \end{enumerate}

Two facts to be explained: \begin{enumerate} \item If someone utters a sentence and you understand her, then you will likely understand others when they utter that sentence. And conversely. \item If a sentence is used to communicate something in one situation, then it can typically be used to communicate much the same thing in another situation. \end{enumerate} An attempted explanation: There are some things and nearly every sentence is related to a different thing. Communicators often know which thing is related to which sentence. This knowledge (is part of what) enables them to understand utterances of those sentences. Terminology: Call these things the ‘meanings’ of the sentences.

facts to be explained

If someone utters a sentence and you understand her, then you will likely understand others when they utter that sentence. And conversely.

If a sentence is used to communicate something in one situation, then it can typically be used to communicate much the same thing in another situation.

What (if anything) are these meaning things?

And what is this relation between the meanings and the sentences?

Consider the first question ... The answer to the first may be mappings from contexts of utterance to truth conditions.

One idea:

The meaning of a sentence is a function from contexts of utterance to propositions.

This should really be complicated : need to go from context of utterance THROUGH context of evaluation TO a truth condition.

Ex: Take ‘I have had breakfast’ and a context of utterance in which Ayesha utters it. The meaning of this sentence takes this context of utterance to the proposition that Ayesha had breakfast.

I won’t attempt to answer this question \citep[see][for an overview of candidate answers to these questions]{speaks:2018_theories}.

‘entities such as meanings ...
are not of independent interest’

\citep[p.~154]{Davidson:1974gh}

Davidson, 1974 p. 154

So now we have a provisional reason to accept that sentences have meanings.

MS, the meaning of the sentence;

PE, the proposition expressed; and

PM, the proposition meant.

Neale, 1990 p. 75

I don’t think there’s any mystery about why we need a PM (actually, often more than one). This is just a matter of saying that when you utter a sentence and communication succeeds, there is something, or some things, which the audience understands.

Why suppose that there is a proposition expressed?

Two facts to be explained: \begin{enumerate} \item \emph{Systematicity} ‘there are definite and predictable patterns among the sentences [utterances of which] we understand’ \citep{Szabo:2004cu} \item \emph{Productivity} communicators can understand utterances of an indefinitely large range of sentences we have never heard before. \end{enumerate} An attempted explanation: Words have meanings, and: \begin{quote} [\emph{Compositionality}] The meaning of a sentence (and of any complex expression) is fully determined by its structure and the meanings of its constituent words. \end{quote}

facts to be explained

Let’s take a look at two further facts to be explained ...

[Systematicity] ‘there are definite and predictable patterns among the sentences [utterances of which] we understand’(Szabó, 2004).

[Productivity] communicators can understand utterances of an indefinitely large range of sentences we have never heard before.

Why distinguish PE and PM? \begin{enumerate} \item (Assumption) MS is a function from contexts of utterances to propositions. \item Suppose for a contradiction that PMs were the values of this function, i.e. MS + context of utterance yields PM. \item Then Compositionality would reqiure a systematic relation between the words uttered and PM. \item (Observation) Substituting words in an utterance can cause PM to vary dramatically, as can what happens after the utterance is over (‘I’ve had a great evening. This wasn’t it’). \item Therefore: what MS + context of utterance yields is not (always) a PM. \item Terminology: Let PE be what MS + context of utterance yields. (Strictly speaking we need context of evaulation too, and further complexities; see \citet{speaks:2018_theories}.) \end{enumerate}

Recall from earler ...

I have had breakfast.

I have had a kidney removed.

I have had fermented fish for breakfast.

I have had a great evening.

Why is this relevant?

Among utterances of these sentences,
there can be is variation in the PM
although Compositionality does not permit corresponding variation in MS.

On the one hand, there can be significant variation in what utterances of these sentences are communicating. That is, the PM varies.

Compositionality does not permit corresponding variation in MS because (a) the structures are similar and (b) the variations cannot be pinned on variations in the meanings of expressions which differ between these sentences. This is most clearly shown by the last two examples, which illustrate that utterances of the same sentence can involve large variations in PM.

Given that MS is a function from contexts of utterance to propositions,
the values of this function will not typically be a PM.

Terminology: Call the value of MS in a given context of utterance the ‘proposition expressed (PE)’.

MS + context of utterance ---yields---> PE

So now we have a provisional reason to accept that there are propositions expresses (PEs) where PEs are distinct from PMs.

MS, the meaning of the sentence;

PE, the proposition expressed; and

PM, the proposition meant.

Neale, 1990 p. 75

What is the relation between PE and PM?

This will be the topic of the next lecture on relevance theory, with special guest lecturer John Michael ...

Let me put the issue another way ...

MS + context of utterance ---yields---> PE

PE + ??? ---yields---> PM

Here is an example of lexical innovation from the master ...

“The Gracehoper was always jigging ajog, hoppy on akkant of his joyicity”

Joyce, Finnegans Wake 408.22

Keyboard Shortcuts?

Lecture 06:

Words and Things

[email protected]

Pragmatic Aspects

Pragmatics vs Donnellan

The Meaning of a Sentence, the Proposition Expressed and the Proposition Meant

Keyboard Shortcuts`?`