Thick Concepts, Implicatures and the Nature of Law

Thick concepts, Implicatures and the Nature of Law1

Abstract: The purpose of the present paper is twofold. First, I

investigate the question whether the evaluative component of

thick concepts is an implicature, a pragmatic enrichment or is

part of the semantics of language. Second, I argue that

debating the evaluations adopted by thick concepts will not be

decisive in a discussion about the correctness of attitudes

usually expressed with the use of a moral predicate. This is

because, the ‘dictionary’ evaluations adopted by those concepts

are conventional assumptions that reflect merely the majority

views and opinions on morality, ethics and normativity in a

society. It is true, that the majority usually adopts a

definite evaluation for some reason. If this reason convinced

so many, then it probably is a reason worth examining.

Nevertheless, this does not entail that it could be a decisive

reason in the debate about ethics or normativity.

a) Introduction

Thick concepts and terms are partly descriptive and partly

evaluative. For example the term or concept ‘courageous’1 This research has been supported by the Polish Ministry of Science and

Higher Education (grant no. DI2012019042).1

describes an action that involves overcoming fear and evaluates

the action as something positive. By contrast, thin terms or

concepts are entirely evaluative, for instance the words ‘good’

or ‘bad’. Nevertheless, this paper is going to focus mostly on

thick terms or concepts. The relation between a thick term and

a thick concept is not obvious. Philosophers often assume that:

‘words have meanings whereas concepts are meanings. (Väyrynen

2012, p. 8) Moreover, they argue that one term may be somehow

linked2 to various concepts (Väyrynen 2012, p. 8) or, that one

concept may be linked to various terms.

While the status of the descriptive components of thick terms

has been thoroughly investigated by philosophers of language,

the character of their evaluative element remains a riddle. The

central question posed by philosophers is whether the character

of the evaluative component is semantic or pragmatic.3 The2 The link usually depends on the theory of meaning assumed. P. Varynen

suggests that concepts may be similar to Fregean senses.

3 This debate is illustrated by Allan Gibbard’s paper and Simon Blackburn’

reply to it entitled ‘Morality and Thick Concepts’. Gibbard probably

endorses the semantic thesis: ‘How do the two components combine? I can

think of three models: that they combine by conjunction, by licensing, or

by presupposition. All three models fail, […], because thick concepts have

too little descriptive meaning to do the work the models demand. Here,2

majority view is that it is pragmatically implicated. I will

try to depict what difficulties this argument faces.

Semantics is roughly the study of meaning. Therefore, if the

evaluative elements were semantic, it would not be dependent on

the context in which the thick term is uttered. By contrast,

briefly, are the models […]. Conjunctive: he conjoins a descriptive

statement and an evaluation. He says that the act is descriptively gopa and

gives a positive evaluation. Licensing: he says that the act is

descriptively gopa. He is licensed by rules of language to use the term

gopa, though, only if he evaluates descriptive gopahood positively.

Presuppositional: his statement presupposes a favourable evaluation of acts

in so far as they are descriptively gopa. It then says that the act is

descriptively gopa. The presuppositions of a statement are the things

hearers must accept for straight agreement or disagreement to be possible.

(Thus for statements that purport to be true or false, they are the things

that must be true if the statement is to be either strictly true or

strictly false.)’ (1992, 274)

By contrast, Simon Blackburn seems to be more of a pragmatist: ‘But this is

left to the […] theory of what a particular speaker is doing on an occasion

by a particular utterance, rather than forged in steel by a prior theory or

convention governing the terms. We might expect someone who talks of a

house as containing south facing windows to be implying or inviting a

favourable attitude to that feature, yet 'contains south facing windows' is

not usually thought of as a thick term, and certainly there is no

linguistic convention that a house with south facing windows should be3

pragmatics is the study of utterances in context. Consequently,

if the evaluative element is pragmatic, the context of uttering

the thick term plays a decisive role in determining the

positive or negative evaluation of the description used.

There are three possible solutions as to the problem posed.

First, the evaluative component is a conversational (pragmatic)

implicature. The difficulties that this view poses will be

presented in the next section. Second, the evaluative element

favourably regarded.’ (p. 287)4

could be merely an impliciture, or pragmatic enrichment.4

Finally, this element could be semantic.

My analysis will be based on several standard assumptions.

First, any natural language is based on a social convention.

Second, meaning will be considered as a set of conditions,4 An impliciture is only a pragmatic expansion of what is said. It cannot be

derived as a conversational implicature. Lawrence R. Horn (2012, p. 21-22)

defines it in the following manner:

‘[…] some aspects of speaker meaning need not be considered either part of

what is implicated or of what is said. Thus consider the following

utterances with the typically conveyed material indicated in curly

brackets:

(16)a. I haven’t had breakfast {today}.

b. John and Mary are married {to each other}.

c. They had a baby and they got married {in that order}. […]

In each case, the bracketed material contributing to what is communicated

cannot be derived as a Gricean implicature, given that it is truth

conditionally relevant, but neither can it be part of what is said, since

it is felicitously cancelable:

(17)a. John and Mary are married, but not to each other.

b. They had a child and got married, but not necessarily in that order.

[…] in such cases the enriched material may be regarded instead as an

impliciture, an implicit weakening, strengthening, or specification of what

is said. This permits an intuitive characterization of propositional

content, a conservative mapping from syntactic structure to what is said,5

which must be satisfied to fulfill the intention to use a term.

(Searle 2009, p. 173) Let me now present the counterargument

for the evaluative component of a thick term being a

conversational implicature.

b) Is the evaluative element a conversational implicature?

c) The truth conditions of what is said are not dependent

upon implicated content.

Consider again the term ‘courageous’. The debate between

cognitivists (those, who think that evaluative statements can

be true or false) and non-cognitivists (those, who think that

we cannot attribute truth values to evaluative sentences) boils

down a decision whether we can say that the proposition:

(I) ‘Being courageous is good’ is true

And

and an orthodox Gricean conception of implicature […]. Bach retains a neo-

classically Gricean semantic characterization of what is said, along with a

post-semantic understanding of conversational implicature: it is

implicItures, not implicAtures that can determine the relevant truth

conditions in such cases. Furthermore, it is misleading to take the

expansions in (16) to be explicatures, since there is nothing explicit

about them, and indeed the cancelability of such expanded understandings

supports their status as implicit..’6

(II) ‘Being courageous is wrong’ is false.

In other words, the content of what is said through uttering a

sentence gives rise to a proposition. A proposition is a

content that is evaluable in terms of truth or falsity. The

content of what is implicated through uttering a sentence can

also be a proposition evaluated in terms of truth and falsity.

However, the content of what is implicated does not influence

the truth conditions of the content of what is said. Consider

the following example:

(III) A: I am out of petrol.

B: There is a garage round the corner. (Grice, 1975,

51)

The answer given by B gives rise to an implicature that ‘the

garage is or at least may be open and has petrol to sell’. This

gives a complete and truth apt proposition. However, even if

the implicature fails and the garage is closed, you still

cannot say:

(IV) It is not true that there is a garage station round

the corner.

7

Therefore, the falsity of what is implicated does not entail

the falsity of what is said. For this reason, viewing value

content as implicated is particularly attractive. (Barker,

2000, p. 269) Regardless of whether you are a cognitivist or

non-cognitivists, you can claim that evaluative content is

implicated. Consequently, even if it is not truth apt, it does

not mingle with the truth conditional proposition uttered. This

way we can account for not truth apt content in language,

without the need to postulate the existence of two different

semantics: one for truth conditional propositions and one for

their evaluative counterparts.5

d) Irony and dual pragmatics

Thick terms are often used ironically in natural language.

Nevertheless, if their evaluative component is a

conversational, pragmatic implicature, then ironic uses of

thick concepts are hard to account for. Consider the following

situation:

5 Hybrid expressivists claim that implicatures express attitudes toward someproperties. They hold non-cognitive views, while having a realistic standpoint toward the existence of moral properties. Nevertheless, they disagree on whether the attitudes are expressed through a conventional or conversational implicature. See (Barker, 2000), (Finlay, 2005), (Fletcher, 2015).

8

(V) A and B are standing in the cellar. B notices a

spider in the corner and starts screaming.

A: You are so courageous!

The context of the utterance as well as the sarcastic tone in

A’s voice indicate that A’s utterance is clearly ironic. Irony

on the classic Gricean account is defined as a conversational

implicature conveying the opposite meaning than the meaning of

the words uttered.6 Let us assume for a moment that the term

‘courageous’ is a mixture of a description and a positive

evaluation in the form of a conversational implicature.

Consequently, the ironic use of the term should convey both an

opposite descriptive and evaluative content. Therefore, the

descriptive content would be the basis for an implicated

opposite, ironic content. By contrast, if the evaluative

component is an implicature itself, then it remains unclear

what could be the character of the opposite, ironic evaluative

content. A meta-implicature? A second-order implicature? The

application of the Occam’s Razor leads us to a much simpler

conclusion: the evaluative component of a thick term is not

necessarily a conversational implicature. 6 This is the Gricean definitione of irony. There are others such as Sperberand Wilson’s account. Nevertheless, as we are using here Gricean notions such as implicatures, the discussed theory of irony will also be Gricean.

9

Ironic use of a thick term conveys an opposite descriptive and

evaluative meaning through an implicature.

standard ironic

description Semantic

meaning

Act,

which

involves

overcomin

g fear

Implicature Act, which

does NOT

involve

overcoming

fear.

evaluation Implicature positive Meta implicature? negative

Thus, if it is not an implicature, it can be either a pragmatic

intrusion or it can be semantic (for instance a conventional

implicature). Consequently, if the first possibility is to be

analyzed, then to explain the ironic uses of the term

‘courageous’ we must resort to the inference scheme proposed by

‘dual pragmatic theories’.

Dual Pragmatic theories state that to achieve propositional

content, we need an amalgam of formal and pragmatic processing.

They remain in opposition to formal theories and pragmatic

theories that deny pragmatic or formal processing respectively.

(Borg, 2004, 3-5) The reason for which formal and pragmatic

theories merged to form dual pragmatic ones was the discovery

10

of ‘unarticulated constituents’ in sentences.7 The basic idea

is the following:

‘U utters ‘s’:

(i) linguistic decoding + pragmatics ⇒ explicature (what

is stated/said)

(ii) what is stated + pragmatics ⇒ what is implied’

(Borg, 2004, 42)

The first step (i) represents the hearer that decodes the

pragmatic, evaluative component of the thick concept. The

second step (ii) represents the hearer that decodes the ironic

implicature on the basis of a thick term with an evaluative

pragmatic element.

The alternative explanation for the ironic uses of thick terms

is the semantic solution. According to this approach, the

ironic conversational implicature is derived on the basis of

the meaning of the words uttered (including the evaluative

component which can be a conventional implicature) plus the

context of the utterance. Consequently, the evaluative, non-

ironic content is a conventional implicature. However, the very7 For a detailed analysis of the subject see F. Recanati „Literal Meaning’

(2004)11

notion of conventional implicature is problematic. For instance

Kent Bach claims that conventional implicatures do not exist.

In fact, they are also some form of a pragmatic linguistic

occurrence. (Bach, 1999)

Nevertheless, to decide between these two possibilities, I must

introduce another typical property of conversational

implicatures: their cancelability.

e) Is the evaluative component cancelable?

It is a necessary, but not sufficient condition for the

existence of a conversational implicature, that it is

cancelable by the speaker.8 Consider the following example:

(VI) A: I am out of petrol.

B: There is a garage round the corner, but I don’t

think it is open. (Grice 1975, 51)

While answering A’s question, B has cancelled the implicature

created by the first part of her sentence, that the garage is

open and sells petrol. Analogously, it may be argued, (Enoch et

8 Grice claimed that conversational implicatures are reinforceable, indeterminate, cancelable and calculable. By contrast conventional implicatures are only detachable. (Fletcher, 2015)

12

al., 2012, 6-7) that the evaluative component of thick terms is

also cancelable, because it is possible to state that:

(VII) ‘Her facing up to her boss was courageous, but it was

not in any way good.’ (Enoch et al., 2012, 7)

The second part of this sentence is supposed to be a

cancellation of the evaluative implicature created by the first

part of this sentence. This argument does not seem convincing

for two reasons. First, if the evaluative content of

‘courageous’ is conversational implicature then it faces the

irony problem. Second, if it is not an implicature, then the

second part of the (VII) sentence is not a cancelation of an

implicature. Consequently, if it is not a cancelation, then we

must answer the question: what is it? There exist many possible

answers. Let us focus on one interesting possibility: the meta-

linguistic statement.

While uttering the first part of (VII) the speaker has made a

choice: to utter a sentence in the English language. The

English language is a convention, like any other natural

language. Therefore, when speaking that language, the speaker

must adhere to its rules, its categorizations and its

13

evaluation. Consequently, the speaker knows that by uttering

‘courageous’ he will be speaking as if he had endorsed the

described act and he will be understood as endorsing it.

Nevertheless, it is perfectly possible that the speaker

personally does not find that acts fulfilling the descriptive

conditions for using the word ‘courageous’ are to be praised.

He may think that the sentence ‘being courageous is good’ is

false and he may be right (depending on whether we adopt the

cognitivist or non-cognitivists approach). Thus, the speaker

can comment on the convention he is using. He can criticize or

discuss it. He can voice his concerns about the evaluative

choices contained in the convention. He may even present

reasons that will convince each and every user of the English

language that the evaluation entrenched in the convention is

wrong and the convention must be changed. However, all of the

above statements are statements in meta-language. They are

statements about the English language. They are linguistically

expressed statements about a linguistic convention. Such

statements often give rise to confusion because natural

languages do not dispose of syntactical markers that would

depict the moment where the switch from natural language to

14

meta-language occurs. Nevertheless, this does not mean that

there is no switch.

Moreover, meta-language enables us to debate both the

descriptive and evaluative conditions for using a term. It is

possible to criticize the term ‘courageous’ not only because it

praises the act called ‘courageous’, but also because it

involves only overcoming fear without for instance overcoming

anger. Therefore, we can equally debate whether the descriptive

conditions for using the term courageous are sufficient,

insufficient, practical, impractical or even inadequate. By

contrast, this meta-debate must be distinguished from a debate

concerning the extension of the word ‘courageous’. It is

possible to argue in English whether some concrete behavior was

courageous or not. In other words, whether it matches the

dictionary description. If it does, then the sentence ‘this

behavior was courageous’ is true. If it does not – the sentence

is false. 9

9 A similar idea has been developed by Stephen Barker, a proponent of

explaining value content through the notion of conventional implicature.

“Disagreement can also be expressed through negation. As ‘T is good’ has

both explicature and implicature, it has two forms of negation: explicature

based, i.e. truth-conditional negation, and implicature based, i.e.15

However, if the description of a thick term is satisfied, then

you cannot debate whether the behavior was good or bad. This is

because the convention does not give you a choice. The

convention can be wrong, but then it needs to be changed. The

positive evaluation of ‘courageous’ is an assumption that can

be changed by convincing all users of the English language in

meta-language that it is wrong and must be altered:

(VIII) If a speaker S:

1. Uses the linguistic convention of English

language.

metalinguistic negation. In reply to someone’s assertion that ‘T is good’, U might

assert:

(17) Jane is not good. She is often heartlessly cruel.

(18)Jane is not good. Being good does not reside in brutal honesty at all

costs.

(17) involves truth conditional negation: the explicature that she is never

heartlessly cruel, which is part of the F-property intended by the speaker,

is the focus of the negation. In (18) the negation is metalinguistic; the

focus of negation is the implicature that a certain F-attitude is shared by

speaker and audience: approval of brutal honesty at all costs.” (Barker,

2000, p. 278)16

2. Speaks about a behavior that fulfills the conditions

for ‚courageous’

Then

1. The convention imposes the practical assumption that

the behavior is praised

2. S is understood by any ordinary user of English as

praising the behavior he describes

One could also wonder what happens if a speaker states that he

has an example of behavior that fulfills the descriptive

conditions for the word ‘courageous’ and the speaker agrees

that generally courageous behavior is good. However, according

to the speaker, this particular courageous behavior is wrong.

Nevertheless, this means that there is something additional

that distinguishes this particular instance of courageous

behavior from other instances. Even if prima facie the behavior

seems ‘courageous’, the ‘all things considered’ judgment is

different, because the term is defeasible. It means that there

is some additional contextual feature, which contradicts the

descriptive conditions of satisfaction. For example someone

behaves courageously because he is forced to do so.

Consequently, we are not disagreeing upon whether to praise

17

someone for such behavior or not. This disagreement is really

about some descriptive elements of the situation. Thus, the

discussion is about the extension of the word ‘courageous’. It

is whether we can qualify that behavior as an example of

courageous behavior in the first place. The decision whether to

praise someone for such behavior or not will be parasitic upon

the decision whether this behavior is courageous or not. If,

despite the contradictory contextual element, we decide to

qualify this behavior as courageous, then the convention will

force us to praise the behavior. If we decide not to qualify it

as courageous, then we can condemn the behavior. As a result,

what the speaker really means by saying that ‘this instance of

courageous behavior is not good’ is that it is no courageous

behavior at all. 10

courageous Debate about the

linguistic convention

(meta)

Debate about the extension

of the word

descriptive whether the

descriptive conditions

whether some concrete

behavior was courageous or

10 We could also interpret (VII) as meaning that the behawior was not good for the individual who faces up the boss because he had to assume some devastating consequences. However, this is not the interpretation that we are intrested here because most of the courageous behaviors are detrimentalto the person who behaves courageously. This is because the predicate ‘courageous’ is moslty associated with dangerous, harmful situations.

18

of satisfaction of

using the term

courageous are

sufficient/practical/a

dequate

not

evaluative whether the evaluation

adopted by the

convention is accurate

(debate about ethics

in meta-language)

AGAIN: whether some concrete

behavior was courageous or not

The fact that the convention of the English language has

adopted the assumption that courageous acts are praised is no

decisive factor in any (meta)ethical discourse. It is just a

practical assumption that facilitates communication.

Nevertheless, it does not entail that courageous behavior is

good and the sentence ‘courageous behavior is good’ is true.

Words are not coined according to universal truths. Quite

contrary, the categories they entrench are formed ‘to the best

of a society’s knowledge’. They are formed so as to serve some

practical, communicative functions. The fact that some

convention adopts one evaluation rather than the other is no

decisive factor in any debate on (meta)ethics. The only thing

19

it can indicate is that statistically speaking when the word

‘courageous’ was invented, most users of the English language

were convinced, for some reason, that courageous behavior

should be praised. Thus, the convention was created. If this

reason convinced so many, then it may be a reason worth

examining in (meta)ethical discourse. By contrast, it cannot be

stated that this is a decisive reason in the meta-ethical

debate. In a nutshell, the idea is this:

1. English is spoken by a large population.

2. The evaluative convention was adopted by so many

for a reason.

3. If this reason convinced so many, then it is worth

examining.

4. However, this does NOT mean, that the choice made

by the convention is a decisive argument in the

ethical debate.

To sum up, if we endorse the semantic thesis (the evaluation is

a conventional implicature), then (VII) can only be read as a

metalinguistic negation of the evaluation of ‘courageous’. A

negation that could be expressed by a non-moralist or an error

theorist. This is a consequence of the fact that a conventional

20

implicature is not cancellable. The problem with this answer is

that we are not sure whether conventional implicatures exists

at all.

However, if we endorse the full pragmatic thesis (it was a

conversational implicature), it could be cancelled because of

the ‘amoralist’:

„who sincerely makes assertions about the moral value of things

but doesn’t subscribe to those moral standards herself and

doesn’t express approval (etc.) by her moral speech acts.

Attitudinal content can be cancelled explicitly, if she merely

explains that she is an amoralist, or that she is contemptuous

or indifferent towards morality. It is contextually cancelled

if her audience already knows of her amoralism (…)” (Finlay,

2005)

Nevertheless, we have established that viewing evaluative

content as conversational implicature is problematic for ironic

utterances. Therefore, a solution could be the claim that

evaluative content is a pragmatic enrichment (an unarticulated

constituent). Then (VII) would not be a cancelation of

implicature but a precisification that the speaker is an

21

amoralist. The problem with this reading is that pragmatic

enrichments do influence the truth conditions of a proposition.

Consequently to adopt this hypothesis we would either have to

be cognitivists or claim that truth conditional semantics

mingle with some different semantics of moral language. The

latter would substantially complicate the picture.

f) The character of the evaluative element

If the evaluative element in thick terms is not a pragmatic

implicature, then it can be either a pragmatic intrusion or it

can be semantic. Nevertheless, there is no need to choose

between those two possible explanations. There are several

reasons to maintain that the evaluative elements in thick terms

have pragmatic roots and semantic outcomes. Therefore, the

evaluative component of a thick concept undergoes, over time,

an evolution – from a pragmatic intrusion toward a fixed or

‘stable’ semantic element. Let us analyze this evolution step

by step in the following subsections.

a) The pragmatic origin

I would like to suggest a hypothesis that seems to be confirmed

by several examples. When a new thick term is coined, it seems

22

to be an objectionable term at first. Some use it to praise the

described behavior, some condemn it and some stay neutral.

Nevertheless, at some point a decision is taken. A decision,

that leads to the formation of a conventional evaluation. A

decision that reflects the majority views on morality in a

society. It is pragmatic in the sense that there usually is a

pragmatic reason, convincing for the majority, which tilts the

balance toward adopting either the positive or the negative

evaluation. Consequently, the evaluative component becomes a

conventional assumption.

Without taking a firm stance in the cognitivist versus non-

cognitivist debate, it is not possible to state that:

(I) ‘Being courageous is good’ is true

And

(II) ‘Being courageous is wrong’ is false

By contrast, it is possible to state that:

(IX) ‘Most people speaking English think that being

courageous is good.’ is true

and

23

(X) ‘Most people speaking English think that being

courageous is wrong.’ is false

Moreover, from the point of view of a theory of communication,

it is possible to state that:

(XI) ‘When you use the word courageous most people

understand that you are praising the act you call

courageous.’ is true

and

(XII) ‘When you use the word courageous most people

understand that you are condemning the act you call

courageous.’ is false.

As a consequence, there is a moment in the thick term

development, when the evaluative component becomes

conventional. It is the moment, when sentences (IX), (X), (XI),

(XII) become true or false respectively. Once the convention is

settled, changing it becomes a challenge. Let us consider an

example from ‘The Boston Globe’ describing the negative

evaluation conveyed by the word ‘marihuana’ and the strive of

activists to shift the term employed to ‘cannabis’, because it

is more neutral:

24

‘Jack Herer’s ‘The Emperor Wears No Clothes,’ a seminal

text for the medical marijuana movement first published

in 1985, lays out a somewhat sensationalised version of

the racist history of prohibition and refers to

cannabis as ‘the plant we denigrate with the slang name

marijuana.’ Since then, and particularly as

legalization battles spread from California in 1996

across the country, ‘marijuana’ has become a

shibboleth. ‘If somebody uses ‘cannabis’ it means he’s more or less

pro-normalization, and someone who uses ‘marijuana’ is anti,’ Mark

A.R. Kleiman, a drug policy expert at UCLA, told me.

When Ricardo Baca became The Denver Post’s first-ever

‘marijuana editor’ last fall, he received a flurry of

e-mails and Reddit messages begging him to change the title

to ‘cannabis editor’ and alter the Post’s style guide

accordingly. He and the Post’s copy chief decided not

to, because marijuana is still the more common term.

But, given activists’ energy, he said, ‘I do think

we’ll see more of the word ‘cannabis’ in the coming

years.’ (Peterson, 2014)

25

The above example illustrates how difficult are the attempts to

change the evaluation associated with a word. The difficulty is

so substantial, that coining a new term with a different

evaluation is much easier. The fact, that the evaluation is so

rigidly grounded in the term invites the conclusion, that the

evaluative component could be context-independent or semantic.

b) The semantic outcome

To test whether the evaluative component of a thick term can be

semantic, let us now address three widely debated examples in

the literature concerning thick concepts.

The first one is the thick term ‘lewd’. Philosophers call it

‘objectionable’ because they find the evaluation of the term

debatable:

‘Some thick concepts are, somehow, objectionable. Somehow

these concepts presuppose or embody values that ought

not really to be endorsed. Gibbard (…) mentions lewd as

an example: he does not agree on the – prude – view on

sexuality which underlies the employment of this

concept. Graham Priest (…) in effect argues that sexually

perverted is an objectionable thick concept, whose usage

26

presupposes that sexual behavior which does not fulfill

a supposed natural purpose is thereby worthy of

condemnation.’(Eklund, 2011, 5)

In my view, the term ‘lewd’ is not objectionable in the sense

that its evaluation has not yet been decided. Quite contrary,

it is a term, which is undergoing the longish and tiresome

process of switching its evaluative component. It is a natural,

sociological and linguistic phenomenon that conventions evolve

and words change their meanings over time. Therefore, ‘lewd’ is

not an example of an objectionable term, because some fifty or

sixty years ago the prevalent view was the prude view. The

fact, that now views are changing and they may slowly outnumber

the conservative ones does not mean that the negative

evaluation is not a conventional assumption. This is because it

has not (yet) been changed. Consider the following sentence:

(XIII) This year’s carnival wasn’t lewd enough.

Through uttering (XIII) the speaker may be trying to suggest

some meta-linguistic conclusion: the evaluation entrenched in

the term ‘lewd’ should be changed. Moreover, consider a

sentence analogous to (VII):

27

(XIV) This behavior is lewd, but I don’t think it is wrong.

Again, the second part of the sentence could be a critical

comment on the convention of English language. A convention

employed by the speaker in the first part of the sentence. As a

result, the term ‘lewd’ may be close to changing its

conventional evaluation, because there are numerous ‘meta

debates’ about it.

A second widely discussed instance of thick concept is the word

‘blasphemous’. It is linked to an anecdote concerning Oscar

Wilde:

‘Oscar Wilde was asked during his trial whether he

denied that something he said or did was an instance of

blasphemy. And Wilde responded: ‘'Blasphemy', sir, is

not one of my words.’ Wilde responded well, for by

either accepting or denying that his act was

blasphemous, he would have committed himself to the

normative judgment underlying uses of the word

‘blasphemous’ – roughly, that speaking against God, the

Church, or some religious tenets and teachings is to be

avoided – and this is precisely the judgment that Wilde

28

was eager to refrain from committing himself to.’(Enoch

et al., 2012, 5)

When uttering the ‘not one of my words response’ Wilde made a meta-

linguistic statement, which adopted a critical stance towards

both the evaluation and the descriptive categorization carried

out by ‘blasphemous’. Therefore, his answer could have been

meta-linguistic. It could have been a statement about the term

‘blasphemy’.

Another debated instance of thick expressions is the term

‘fashionable’. Its evaluative element is aesthetic rather than

ethical. Nevertheless, it carries out a positive conventional

assumption.

‘(…)we can think of judgments about what is and what is

not fashionable made by a participant (for whom

proclaiming a suit fashionable amounts to some kind of

endorsement), and also by a fashion-historian (who

refrains from endorsing any standards of

fashionableness).’ (Enoch et al., 2012, 14)

What could be problematic in the above quotation is the

attitude of the fashion-historian. This is because she decides

29

to use the convention and employ the term ‘fashionable’.

Therefore, she utters the word ‘fashionable’ in English. Yet it

is clear that she does not commit herself to the positive

evaluation conveyed by the term. As a consequence, she utters

the word AS IF she endorsed the positive evaluation carried out

by the term. Therefore, it is possible for the speaker of any

language to decide to use that language, while not endorsing

the evaluations that it entails. The speaker acts as if he

endorsed the evaluation by uttering the word. She may have some

practical reasons to do so. Moreover, the utterer is aware that

she will be understood as endorsing the view. This

understanding of the speaker can be cancelled by the context of

her utterance. For instance, if she utters it as a fashion

historian, then the hearer will usually infer that it is not a

personal commitment of the speaking person. As a result, she

acts as an aesthetic amoralist.

Words or sentences uttered in a specific context can change

their meanings. According to Paul Grice, such is the difference

between sentence meaning and speaker’s meaning. Consequently,

this applies also to thick terms. As a result, there exist

contexts in which the speaker will be understood as conveying a

30

different evaluation, than the conventional one. For example,

the term ‘communist’ may be differently perceived when uttered

in a post-soviet society, than during the meeting of a left-

oriented political party. However, this does not alter the

fact, that there is some conventional, dictionary evaluation

that must be changed by the context of utterance.

In conclusion, the status of the evaluative element in thick

terms could be a practical assumption, determined by the

convention used. Nevertheless, this assumption becomes settled

only when a majority decision is taken. This decision often has

a practical, pragmatic reason behind it. As a result, the thick

concept has pragmatic roots and a semantic outcome formed over

time. The convention is not settled once and for all. It can

evolve if the community decides to change it. Yet this requires

another decision backed up by the majority of the ordinary

users of a language. As we have tried to establish a general

view of thick concepts, let us now turn to a pertinent question

for lawyers: what are the consequences of considering the

normative element in the term ‘legal’ as semantic?

(I) The normative component of the term legal

31

The consideration in this section will be made on the basis of

the following hypothesis:

‘We suggest, then, that ‘legal’ be thought of as a

thick term, and the concept legal as a thick concept. The

concept’s descriptive content can then be understood in

terms of representations of some social facts -- i.e.

the social facts in virtue of which some act or

practice type counts as legal or illegal. These facts

would differ from one jurisdiction to the next, but

they may be uniform across jurisdictions on

sufficiently high levels of abstraction. But as with

other thick concepts, that it has descriptive content

does not preclude its being evaluative as well. Indeed,

often, or perhaps even necessarily, declaring an act

legal (or illegal) would involve an expression of some

evaluative or normative commitment. There may be an

underlying normative judgment involved here – perhaps

something to the effect that the fact that an act

satisfies the descriptive criteria for legality is a

reason for certain officials to permit them, or perhaps

to the effect that the fact that an act fails to

32

satisfy these descriptive criteria counts strongly

against these officials permitting it.’ (Enoch et al.,

2012, 11)

Thus, the term ‘legal’ could also be composed of a descriptive

plus normative component. In their paper, D. Enoch and K. Toh

do not commit themselves to a decision whether the normative

component is semantic or pragmatic. Therefore, the question

arises: does the semantic character of the normative component

proposed in this paper entail any particular consequences? Let

us define ‘normative’ for the purposes of this paper as giving

reasons to act in accordance with the law.

The Oxford dictionary defines the term ‘legal’ as:

‘Relating to the law, appointed or required by the law,

recognised by common or statute law, as distinct from

equity, permitted by law’ (Oxford Dictionaries)

Therefore, if an element satisfies the descriptive conditions,

then it is legal. Consequently, according to the convention,

the speaker is given a reason to behave in a way that satisfies

the description. Speaking in Hartian terms, the utterer can be

either adopting the internal point of view – then he really

33

thinks that the act he calls ‘legal’ gives him one of the

reasons he has to act that way, or the external point of view –

then by using the word ‘legal’ he pretends or presupposes that

it gives him a reason to act that way. If a speaker decides to

state that some behavior is legal, then he acts AS IF he had a

reason to behave this way. Moreover, such speaker will be

understood as having a reason to follow the behavior he calls

legal:

(II) If a speaker S:

1. Uses English language.

2. Speaks about a behavior that fulfills the conditions of

satisfaction of ‚legal’

Then

1. The convention imposes the practical assumption that S

has a reason to act that way

2. S is understood by any ordinary user of English as

having a reason to act that way

Again an ordinary user of the English language may state that:

(XV) This behavior is legal, but I do not think you have a

reason to act this way.

34

Again, this sentence can be interpreted twofold. First, the

speaker may be pointing out to some additional contextual

features of the particular situation that contradict the

descriptive conditions of satisfaction of the word. For

example, the behavior may infringe some soft law or

international law principles that are not directly binding upon

the individual. Consequently, the debate is in fact a debate

about the extension of the word ‘legal’. It is a discussion

whether the behavior is an example of legal behavior or not.

The decision whether the individual has a reason to act in the

discussed way is parasitic upon the decision whether the

behavior satisfies the dictionary description of ‘legal’. The

individual will truly have the reason only if he adopts the

Hartian internal point of view. As a result, the debate is not

about whether the individual has a reason to act or not, but

about whether the behavior is legal or not.

Second, if there are no contextual features that could

undermine the fulfillment of the descriptive conditions of

satisfaction of ‘legal’, then the second part of (XV) is a

sentence in meta-language. It is a criticism of the English

language. It debates whether the fact that a behavior falls

35

under the scope of the dictionary definition of ‘legal’ is a

sufficient reason to act this way. It discusses the practical

assumption made by the convention: that if a speaker can

describe some behavior as legal then he has a reason to behave

in that way. As a result, only this meta-linguistic debate will

be a true debate about normativity, free from any conventional

assumptions. Assumptions that may be practical but not true in

the cognitivist sense. Stating that some behavior is truly

normative because the term ‘legal’ applies to it is a weak

ethical argument.

legal Debate about the

linguistic convention

(meta)

Debate about the

extension of the word

descriptive whether the

descriptive conditions

of satisfaction of

using the term legal

are

sufficient/practical/a

dequate

whether some concrete

behavior was legal or

not

evaluative whether the

fulfillment of

AGAIN: whether some

concrete behavior was legal

36

conditions of

satisfaction gives you

a reason to act that

way

(debate about

normativity in meta

language)

or not

In a nutshell, the argument proceeds as follows:

1. A large population speaks English.

2. The normative convention was adopted by so many for a

reason.

3. If this reason convinced so many, then it is worth

examining.

4. However, this does NOT mean, that the choice made by

the convention is a decisive argument in the debate about

normativity.

The debate over the extension of a word and over a linguistic

convention again needs not to be confused.

5. Conclusion

This paper is an attempt to evaluate the pros and cons of

viewing the evaluative component of thick concepts as a

37

conversational implicature, pragmatic enrichment or semantic

element. Nevertheless, the main idea is that debating the

evaluations adopted by thick concepts will not be decisive in a

discussion about ethics, meta-ethics or normativity. This is

because; the evaluations adopted by those concepts are only

practical, conventional assumptions that reflect merely the

majority views and opinions on morality and ethics in a

society. It is true, that the majority usually adopts a

definite evaluation for some reason. If this reason convinced

so many, then it probably is a reason worth examining.

Nevertheless, this does not entail that it could be a decisive

reason in the debate about ethics or normativity. Finally,

while D. Enoch and K. Toh in their paper do not provide

decisive arguments as to whether the evaluative component of

the term legal is semantic or pragmatic, this paper argues that

it has pragmatic roots and a semantic outcome.

References:

1. Bach, K., 1999. The Myth of Conventional Implicature.

Linguist. Philos. 22, 327–366.

38

2. Barker, S., 2000. Is value content a component of

conventional implicature? Analysis 60, 268–279.

doi:10.1093/analys/60.3.268

3. Blackburn, Simon, Gibbard, Allan, 1992, Morality and Thick

Concepts, Proceedings of the Aristotelian Society,

Supplementary Volumes, Vol. 66, pp. 267-283+285-299

4. Borg, Emma, 2004, Minimal Semantics, Oxford University

Press

5. Eklund, Matti, 2011, What are thick concepts?, Canadian

Journal of Philosophy, Volume 41, Issue 1

6. Enoch, David and Toh, Kevin, 2012, Legal as a Thick

Concept: W.J. Waluchow & Stefan Sciaraffa eds., The Nature

of Law: Contemporary Perspectives, Oxford University Press

7. Finlay, S., 2005. Value and Implicature. Ann Arbor MI

Mich. Publ. Univ. Mich. Libr. July 2005 5, 1–20.

8. Fletcher, G., 2015. Moral Utterances, Attitude Expression

and Implicature, in: Fletcher, G., Ridge, M. (Eds.),

Having It Both Ways: Hybrid Theories and Modern

Metaethics, Oxford Moral Theory. Oxford University Press,

Oxford ; New York.

9. Grice, Herbert P., 1975, Logic and conversation: P. Cole &

39

J. Morgan (ed.), Syntax and Semantics, 3: Speech Acts: 41–

58

10. Horn, Lawrence R., 2004, Implicature: Handbook of

Pragmatics, ed. Horn L. R. Ward G., Blackwell Publishing

pp. 21-22

11. Oxford Dictionaries

[http://www.oxforddictionaries.com/definition/english/lega

l accessed on 09.09.2014]

12. Peterson, Britt, 2014, This is your language on

cannabis As marijuana goes legit, our words for it are

changing, too, The Boston Globe,

[http://www.bostonglobe.com/ideas/2014/01/26/this-your-

language-cannabis/9bguslv7ZkSsHTElfANTTN/story.html

accessed on 26.01.2014]

13. Recanati, Francois, 2004, Literal Meaning, Cambridge

University Press

14. Searle, John R., 2009, What is Language? Some

Preliminary Remarks: Etica & Politica / Ethics & Politics,

XI (2009) 1, pp. 173-202

40

15. Väyrynen, Pekka, 2012, Thick Concepts: Where’s

Evaluation?: Russ Shafer-Landau (ed.), Oxford Studies in

Metaethics, Volume 7, Oxford University Press. 235-70

41