To What Extent Do Proper Names Resemble Indexicals

To What Extent Do Proper Names Resemble Indexicals?

Paula Milne

Paula Milne (pm34@st-andrews.ac.uk) is a Senior Honours Philosophy student at the University of St Andrews. Her main areas of interest are Philosophy of Language, Philosophy of Mind and Epistemology. After graduating, Paula will stay at St Andrews for a masters course.

Introduction

The purpose of this paper is to consider to what extent proper names can be said to resemble indexicals and in particular to consider a recent claim by Michael Pelczar that names are indexicals. The paper focuses on work by Francois Recanati^[1], John Perry^[2] and Michael Pelczar^[3]; three philosophers who hold a 'indexical' theory of proper names but who compare proper names and indexicals to varying degrees. The particular focus of the paper is the debate between John Perry and Michael Pelczar regarding the indexical nature of proper names.

Perry, Recanati and Pelczar offer variations of an indexical theory of proper names; a theory which claims to solve some of the problems associated with standard direct reference theories. As the indexical theory is fairly recent this paper presupposes no knowledge of the theory. Section 1 gives a short introduction to a standard direct reference theory of names, a brief account of a standard theory of indexicals, a summary of an indexical theory of proper names and finally offers some motivation for the indexical theory of names. It should be understood that the information in Section 1 is given as essential background for the purpose of the paper; it is not the purpose of this paper to defend or criticise the indexical theory of names as a whole, nor to consider critically any theory of indexicals or proper names. Section 2 focuses more closely on the views of Recanati, Perry and Pelczar, giving a summary of the area of agreement between the three philosophers before focussing on the crucial differences between their theories. Finally, in section 3 of the paper we are in a position to consider whether proper names can be said to resemble indexicals or whether in fact they should be included in the class of indexicals.

1 Proper Names, Indexical, and an Indexical View of Proper Names

1.1 Proper Names - A very short introduction

Consider the following two quotes concerning proper names and direct reference;

Proper names are often considered the prototype of directly referential expressions (Recanati)^[4]

Names have been thought to challenge direct reference theory (Kaplan)^[5]

These quotes appear to be conflicting, yet they are both true. Since Kripke's attack of descriptivism^[6], direct reference theories of various kinds have been prominent and as Recanati's quote states, proper names are often considered prototypical of this kind of term. A directly referential term is one that's role is simply to refer, contributing an object to the proposition in which it occurs. Because of this it can be persuasively argued that a proper name has no 'meaning'; if asked what the term 'Tony Blair' meant many people would reply that it doesn't mean anything, it simply 'stands for' a person who is named 'Tony Blair'. Unlike nouns such as 'lamp', there is no criteria which must be fulfilled in order for an object or person to be named 'Tony Blair', or any other name for that matter; the name is simply to be understood as a means of picking out a particular person, thing or place.

However, as the quote from Kaplan suggests, many have been left dissatisfied with direct reference theory because of its inability to adequately explain its own 'paradigm' example; the proper name. It seems that names do present more than just the reference to the proposition as shown by this classic example of a substitutional puzzle. If a proper name is directly referential and it contributes only its reference to the proposition, then two terms with the same reference should be interchangeable without any loss of significance. Therefore, given that 'Hesperus' and 'Phosphorus' are two names inadvertently given to the same planet the following propositions:

(1) Hesperus is Phosphorus, and

(2) Hesperus is Hesperus

must be identical. Yet (1) is informative while (2) is trivial, therefore the two propositions cannot be identical therefore proper names must contribute more than just their reference. This is because the substitution of one term for another that has the same reference can change the impact that the proposition has on the listener or reader. Therefore, the challenge is to find a theory which is directly referential in that it captures our intuitions regarding names being 'rigid designators'^[7] yet can account for the obvious fact that names often contribute more than just their referents. The indexical view of proper names claims to fulfil these needs.

Before presenting the theory it will be beneficial to consider a brief account of a theory of standard indexicals in order to understand what comparisons can be made between the two types of terms.

1.2 Indexicals

Indexicals are complex terms but in this paper it is only possible to present what is taken to be an uncontroversial theory of indexicals in order to make a comparison with proper names. In 'Demonstratives'^[8] Kaplan gives the standard list of indexicals as; personal pronouns 'I', 'you', 'my', 'he', 'his', 'she' and 'it'; demonstrative pronouns 'that' and 'this'; adverbs 'here', 'now', 'today' and 'tomorrow'; adjectives 'actual' and 'present'.^[9] Indexicals are directly referential terms in that they contribute the referent to the proposition and not just a mode of presentation of the referent. However the meaning of an indexical cannot be simply the referent as each token of the same indexical type can have a different referent. For example, if I say 'I' this token refers to me but if you say 'I' the token will refer to you, therefore we must distinguish between the 'character' of an indexical and the 'content' of an indexical^[10]. The basic character/content distinction is best understood as follows; due to the character (meaning or rule) of the indexical type, the referent (content) will be presented in a certain way. We can consider the indexical 'I' as an example. The character of 'I' can be understood as a linguistic rule such as;

P: In each of its occurrences, 'I' refers to the person who is speaking or writing^[11]

It is this character which can be considered as the meaning of the indexical term; so for indexicals, meaning is to be considered as a property of expression types rather than a property of a particular utterance. Therefore, the meaning of an indexical is fixed by the conventions we learn when we learn a language, not by its particular uses.^[12] Following from this meaning, at each particular use of 'I' the rule given at (P) will fix the referent dependant on the context of utterance; the referent is the 'content' of the token. We understand what the content of a particular 'I' is because the character of the indexical takes us from context to content.^[13] Therefore, unlike the character of 'I', the content is a property of the individual utterance.

This distinction between character and content leads to a further distinction in the truth conditions of a statement containing an indexical. We can distinguish between reflexive truth conditions and, to use Perry's terminology, 'incremental' truth conditions.^[14] The reflexive truth conditions are purely linguistic, about the utterance itself. Supposing (P) to be an adequate linguistic rule for 'I', a statement of the form

(3) 'I like bananas'

will have the following reflexive truth conditions

Q: There is an x such that i) x is the speaker of (3); ii) x likes bananas.

If (3) is heard as a cry from a room full of people where the speaker can not be identified it can still be understood and the truth conditions for the utterance can be given as (Q). There is of course further knowledge available, knowledge of who the speaker is. If it is discovered that Tony Blair is the person responsible for (3) then 'x' can be replaced with 'Tony Blair'. Then the truth conditions for (3) can be stated as follows

R: Utterance (3) is true iff Tony Blair likes bananas.

Perry states that the context dependence of indexicals is often taken to be their defining feature as what an indexical designates shifts from context to context. However, there are different possible applications of 'context' and in order to individuate indexicals we need to be clear about what precise role context plays. This is considered in more detail in section 3.3.

1.3 The Indexical View of Proper Names^[15]

An indexical theory of proper names incorporates elements of descriptivism with a form of direct reference theory. It achieves this by employing a distinction similar to the character/content distinction given for indexicals in section I.2. Recall that although the meaning of the indexical is said to be given by the character there is also an obvious sense in which the content could be considered to be the meaning of an individual occurrence. Similarly, a proper name can be considered as having two different types of 'meaning'; a linguistic meaning which gives all names a fixed content regardless of context, and a non-linguistic^[16] meaning which links the name token to a social ‘naming’ convention. At the simplest level the linguistic meaning looks like any other metalinguistic theory of proper names;

R: By virtue of its meaning, a proper name NN indicates that there is an entity x such that an utterance S(NN) is true iff x satisfies S( ). ^[17]

Proper names are still 'rigid designators', the rigidity arising directly from facts about lexical meaning.^[18] However there must be more to the meaning of a proper name if we are to concede the obvious fact that proper names are used as a fairly successful means of directing our listener to a particular object; there must be a further explanation for the fact that the term 'Tony Blair' can pick out a particular individual. This further 'meaning' is not a linguistic convention but a social^[19] convention. It is this social convention which is established when a thing or person is assigned a name. When Tony Blair was named 'Tony', a convention was established that made it possible for people to designate him with the name 'Tony Blair'. It must be emphasised that this social naming convention is not part of the linguistic meaning of 'Tony Blair'. The linguistic meaning is almost exhausted by (R); almost as, although the naming conventions are not linguistic, the fact that each proper name has an associated naming convention must be treated as a linguistic fact. Therefore (R) must be amended to include the fact that a proper name has (or at least is intended to have) a bearer. Linguistic competence concerning proper names does not require us to know who the bearer is but it does require us to know that a name is supposed to have a bearer. Note again that although the linguistic meaning now makes reference to the social convention, the social convention itself is not part of the linguistic meaning. The complete linguistic meaning of a proper name can be stated as follows;

R*: By virtue of its meaning, a proper name NN indicates that there is an entity x such that an utterance S(NN) is true iff x satisfies S( ). For each type of a proper name there exists (in principle) a social convention linking that name to an individual.

For example, the following sentence

(4) Tony Blair likes porridge for breakfast.

is true iff there is an x who is the bearer of the name 'Tony Blair' and (4) is exploiting this naming convention, and this x likes porridge for breakfast. Notice that by this definition the truth of (4) in no way requires acquaintance with Tony Blair; the truth conditions depend purely on linguistic facts concerning the linguistic meaning of a proper name. It is now possible to distinguish between two levels of understanding. In order to have linguistic competence and understand (4), a user of the language must simply be acquainted with (R*), however full understanding requires us to have access to the social convention that associates this particular token of 'Tony Blair' with an individual.

A further consequence of this theory which becomes relevant later in this paper is that the indexical theory treats two names of the same form e.g. two 'Paula Milnes', as being ambiguous and not homonymous. A homonymous view would say that the terms are individuated by meaning i.e. a description for descriptivists or perhaps speakers intention as in Kaplan's theory. The indexical theory can not individuate these two by meaning as they both mean 'the bearer of 'Paula Milne''. Therefore, the name is ambiguous. In fact most names are multiply ambiguous as many people share the same name type (or form) so there will be many associated conventions linking the name to many different referents. The context (the salient 'Paula Milne' convention in that context) serves to disambiguate and, as Perry says, if in doubt you can always just ask, 'Which Paula Milne are you talking about?' The ambiguity of names is considered in more detail in Section 3.

1.4 Motivation for the Indexical View of Proper Names^[20]

The indexical theory of proper names claims to provide solutions to many of the problems associated with other direct reference theories for proper names. For example Gareth Evans^[21] famously criticised Kripke's direct reference theory because of its inability to deal with 'shifting' reference. Kripke's theory of referring (the causal theory) states that all subsequent uses of a name are causally linked to an initial baptism and that the object present at the initial baptism is the referent of the term. Evans uses the example of 'Madagascar'; a name originally given to a part of the African mainland which, through a misunderstanding came to be used to refer to the island now known as Madagascar. Kripke's theory forces reference to the part of the African mainland when the intended referent is the island. This is because it was the part of the mainland which was involved in the original baptism. The indexical theory claims to solve this problem as the theory of referring involves social conventions, conventions which are established within the context of the current conversation. The current dominant social convention associated with 'Madagascar' links it to the island and someone who wanted to use the term to refer to the part of the mainland would have to specify the context by saying something like, 'the part of the mainland which was previously referred to as 'Madagascar'. The theory can also account for reference at the time of uncertainty when the name 'Madagascar' was being used by some people to refer to the mainland and others to refer to the island. According to the indexical theory there were two social conventions linked to the same term therefore 'Madagascar' was an ambiguous term, relying on context to differentiate each particular use.

In section 1.1 it was shown that direct reference theories have difficulty explaining a possible difference in cognitive attitude when asserting

(5) Hesperus is Hesperus, and

(6) Hesperus is Phosphorus

when 'Hesperus' and 'Phosphorus' are names for the same object. For example, someone who was not aware that these two terms named the same object would assert (5) while denying (6). The indexical view can explain how a competent user of the language can have different attitudes regarding 'Hesperus' and 'Phosphorus' while agreeing that, as rigid designators, these two terms may contribute the same object to a proposition. This is because linguistic competence does not require us to be acquainted with the social naming conventions that are being exploited. An individual can have incompatible attitudes towards the sentences 'Hesperus is shining' and 'Phosphorus is shining' in the same way that an individual can have incompatible attitudes towards 'A bear is about to attack him' and 'A bear is about to attack me' when (unknown to the speaker) 'him' and 'me' pick out the same individual.^[22]

To summarise section 1, it appears beneficial to understand proper names as bearing a linguistic convention and a social convention, hereby avoiding the standard problems associated with direct reference theories while retain the desirable quality of names being rigid designators. Section 2 focuses more closely on the views of Recanati, Perry and Pelczar; three philosophers who make comparisons between proper names and indexicals but differ in the extent to which they want to stress the comparison.

2. Three Theories Comparing Proper Names with Indexicals

The purpose of this section is to focus on the subtleties of the theories presented by Recanati, Perry and Pelczar. The section begins with a summary of the elements of indexicality which the three philosophers would all concede as being comparable with proper names, before considering differences between the theories.

2.1 The Area of Agreement

In the theory detailed in section 1.3 proper names have much in common with indexicals. Consider the following facts. The linguistic meaning or reflexive content of an indexical, an example of which is (P), refers the hearer to a relation holding in context between the expression and the reference. This means that a hearer of (3) who understands English would know the truth conditions required to make the statement true, without requiring the hearer to know who the speaker is. Similarly the linguistic meaning or reflexive content of a proper name, as shown in (R*) allows a hearer who is competent in English to understand a sentence such as (4), without requiring the hearer to have knowledge of the referent of the proper name expressed. In both cases there is further knowledge available to the hearer, knowledge which depends in some way on the context of the utterance. Each indexical type has a fixed linguistic meaning, a rule or character, which takes the hearer from context to content. Proper names also have a fixed linguistic meaning which refers the hearer to a relation (the name bearer relation) holding in context between the name and its referent.

Despite this communal start point Recanati, Perry and Pelczar have differing attitudes regarding the indexicality of proper names.

2.2 Recanati - The 'Indexical View'

Recanati claims that 'proper names are very much like indexicals'^[23]. He sees proper names as a 'special case'^[24] of indexicals and it is perhaps this 'specialness' that allows him to concede that there are differences between proper names and standard indexicals. Firstly, indexicals are 'token-reflexive'; a token of 'I' refers to the utterer of this particular token, not to general utterances of 'I'. Conversely, proper names are 'type-reflexive'. The relation (name bearer relation) which holds between a token of NN and its bearer is a relation determined by the name type. Recanati holds that this is not a crucial distinction and points out that proper names can be considered to be 'token-reflexive'. This is because, although the main relation is 'type-reflexive', it is a particular tokening of the name which operates in each instance; they are 'token-reflexive' in terms of the context of this utterance. A further difference which Recanati concedes is that, while each type of indexical specifies a different relation (character), each proper name has the same relation; there is a 'single semantic rule'^[25] for every proper name. Therefore, all proper names have the same linguistic meaning, while each type of indexical has a different linguistic meaning. Recanati believes that this is a notable difference, therefore he does not go so far as to say that proper names are indexicals;

For indexicals are individuated partly by their meaning, while proper names all have the same meaning and are individuated only by their form^[26]

Overall Recanati appears uncertain about the extent to which he wants to draw the analogy. He does call his theory of proper names 'The Indexical View' and seems happy to refer to proper names as a 'special case of indexicals' yet, despite this, Recanati retains the distinction between proper names and indexicals.

2.3 Perry - Similar but Different

In Reflexivity, Indexicality and Names, John Perry describes proper names and indexicals in much the same way as Recanati, yet he is not willing to say that proper names are 'indexical-like'. For Perry there is an important difference between proper names and indexicals which make it impossible to acknowledge more than an interesting similarity between the two. The distinction centres around the role of context. Essentially, although context plays a role in determining the reference for both indexicals and proper names, it plays a very different role in each of the two cases. With indexicals it is the meaning of a particular expression which determines that a certain relationship will hold and determine the reference; the meaning of the indexical type determines the contextual relation. The role of context regarding proper names is very different; the role of context here is to make a potentially ambiguous term less ambiguous. In Perry’s own words:

Many names have thousands of meanings - that is, there are thousands of individuals that they are used to designate, exploiting various permissive conventions.^[27]

The role of context helps to 'narrow down the possibilities for the conventions that are being exploited'.^[28] Perry makes his position very clear when he states that, with names, we have ambiguity, not indexicality.^[29] This is because the contextual element given by the rule of a proper name is very different to the contextual element dictated by the rule for an indexical. This important distinction will be considered in more detail in Section 3.3.

2.4 Pelczar - Proper Names are Indexicals

In Names as Tokens and Names as Tools^[30], Michael Pelczar argues that proper names are indexicals. In order to make this claim, Pelczar proposes that naming conventions, which he calls 'dubbings-in-force'^[31] are contextual factors in the same way as the standard contextual factors; utterer, time, spatial position and possible world.^[32] Pelczar claims,

Taking dubbings-in-force as a feature of contexts of utterance results in the assimilation of all expressions that are introduced in dubbings - including all (typical) proper names - to the class of indexicals.^[33]

By including dubbings-in-force as a contextual factor Pelczar argues, contra Perry, that names are indexical, not ambiguous. For example, a view such as Perry's would treat the name 'George Bush' as having a character which assigns to each context of utterance, a function from possible worlds to George Bush, but the name is lexically ambiguous and in formal semantics a subscript would be introduced to distinguish between the different bearers of the name. Pelczar argues that by introducing dubbings-in-force as a contextual factor names are no longer ambiguous; the dubbings-in-force relevant to the context of utterance enable the name to be treated as a 'rigidly designating indexical'^[34]. So, just as the standard indexical 'I' is sensitive to the 'utterer' feature of a context of utterance, 'George' is sensitive to the 'dubbings-in-force' feature of a context of utterance.^[35] ^[36]

To summarise section 2, it is clear that, while the views of the three philosophers considered share some common ground they differ in the extent to which they want to compare names and indexicals. However, Recanati and Perry's theories turn out to be remarkably similar. The two theories appear to be distinct for two reasons. Firstly there are terminological difference. However, it seems that Recanati's linguistic and non-linguistic conventions can be equated to Perry's reflexive and incremental distinction. Secondly, as discussed in this section, the two philosophers differ in the extent to which they would want to say that names are like indexicals. However, this is purely a difference in attitude towards the comparison rather than a disagreement over common factors between names and indexicals; Perry has been more proactive in pointing out the differences between the two terms while Recanati has been content with the similarities. Therefore, it is now concluded that Perry and Recanati are offering the same basic theory and as such it makes sense to discuss them as one theory from this point onwards and compare this theory with Pelczar's. Given this, it will prove beneficial to refer to Perry and Recanati's theory under a title which distinguishes it from Pelczar's theory. It is misleading that Recanati calls his theory an 'Indexical View' as he himself agrees that names do not behave like indexicals; the two classes of terms simply share an element of reflexivity. Therefore, it would clarify matters if the theory of proper names shared by Recanati and Perry be referred to as The Reflexive Theory of Proper Names while Pelczar's strong theory be referred to as the Indexical Theory of Proper Names. The rest of this paper will refer to the theories in this way.

3 To What Extent do Names Resemble Indexicals?

This final section focuses on two papers; 'Reflexivity, Indexicality and Names' by Perry and 'Names as Tokens and Names as Tools' by Pelczar. Perry proposes what we can now refer to as the Reflexive Theory of Proper Names and gives reasons why names can not be indexicals. Pelczar's paper directly addresses Perry's paper and argues that Perry has given no good reason why names can not be indexicals before concluding that names are indexicals. The purpose of this section is to determine whether names resemble indexicals or whether names are indexicals. As yet Perry has not replied to Pelczar's challenge (at least not in print) therefore, both the defence of Perry's position and the analysis of Pelczar's argument which appear in this section are original to this paper.

3.1 Pelczar's criticism of the Reflexive Theory of Proper Names

Pelczar's criticisms divide into two parts in his paper; a general argument directed at the claim that names are ambiguous, then an attack on Perry's argument against names being indexicals.

3.1.1 Pelczar's 'Direct Argument' Against the Ambiguity of Proper Names

Pelczar offers a 'direct argument' against Perry's claims that names are ambiguous, not indexical. He argues that, as there are important differences between standard ambiguous terms (like 'bill'^[37]) and names (like 'Bill'), names cannot be ambiguous. According to Pelczar, not knowing one of the meanings of an ambiguous term such as 'bill' does not stop a person being a competent user of the term. For example, if someone did not know that 'bill' could be used to refer to a narrow strip of land, they may well be confused by

(5) The tractor drove across the bill.

This would not, however, affect their linguistic competence concerning the other meanings of 'bill'; for example, they may well still understand

(6) I received a huge gas bill.

Pelczar's point is, this would not happen if someone did not understand a sentence using an 'ambiguous' name such as 'Bill'. If they understood all of the other terms in the sentence but not the naming term, we would say that this one instance was enough for us to claim that this person was not a competent user of the term 'Bill'.^[38]

3.1.2 Pelczar's response to Perry's Argument

Pelczar then responds to Perry's arguments against the indexicality of proper names with the aim of concluding that proper names should be included among the class of indexicals. Pelczar locates just one argument in Perry's paper against the indexicality of names and considers various construals of this argument. A succinct account of the argument is given by Pelczar as: 'the literal content of an indexical varies in accordance with "certain specific contextual relationships" determined by its meaning; the content variability of a name is not of this regimented sort; hence, names are not indexicals.'^[39] In an effort to get to the heart of Perry's objections Pelczar considers three possible interpretations; (a) Lack of Specificity, (b) Messiness and (c) Underdetermination.

(a) Lack of Specificity: Pelczar suggests that Perry's worry could be as follows; whereas each indexical has a specific rule peculiar only to it, the rule for a proper name lacks specificity; it corresponds to an entire part of speech (the semantic category of names). Pelczar agrees that this is the case for some indexicals, for example there is a specific rule attached to 'yesterday' and a different rule for 'now'. However, if Perry's point is that each indexical has a specific rule which applies to it alone (unlike proper names which have one rule for the whole category), the indexicals 'this' and 'that' appear to be a counterexample as the rules for these two indexicals prove to be very difficult to individuate.

(b) Messiness: Contextual relations determining content for proper names are extremely messy and complex (simply gestured at by generalisation such as 'social convention' and 'name bearer' relation). Indexicals, on the other hand, determine content in a relatively straightforward manner. Pelczar agrees that contextual determination of a given name can be extremely complicated but he argues that it is not clear that all indexicals lack this complication. He distinguishes between 'tidy' indexicals such as 'yesterday' and 'messy' indexicals such as 'here'. The content of 'here' can be described as the 'function for evaluation points to the location in which that context takes place'^[40]. Pelczar argues that the use of 'location' is hiding the complexity of this indexical, rather like the gesture to a 'name bearer' relation can be said to hide the complexity of the reference determination of a proper name. For example, if I say 'I live here', the only factor which will distinguish between

(i) 'I live in this room'

(ii) 'I live in this house'

(iii) 'I live in this village'

(iv) 'I live in this country' and so on

will be an appeal to the content of the proposition I am intending to refer to. However, Pelczar argues, there is no reason to suppose that speaker's intention should override all the other contextual factors that are considered when determining a complete content for 'here'.^[41] Pelczar's point is this; no matter how we try to supplement the rule for 'here' the 'tidiness' of this indexical will always be an illusion. Therefore, the fact that proper names are 'messy' is not going to be sufficient to exclude them as a category of indexicals.

(c) Underdetermination: In the context of which a name is uttered the rule for a particular name leaves the content underdetermined, whereas each indexical rule fully determines the content with respect to the context of the utterance. Again, Pelczar concedes that indexicals such as 'yesterday' do not require any element of 'speaker discretion'; there is no ambiguity regarding the content of a given token. However, this is not the case with all indexicals. For example, an occurrence of 'that' requires some appeal to speaker's intentions as there are many things which could be taken to be the content of each instance; this element must be included in the linguistic rule therefore it is clear that the content of the indexical, as with proper names, is underdetermined by the linguistic rule.^[42]

3.2 Analysis of Pelczar's criticism

In this section it is argued that Pelczar's criticisms of Perry's paper are not successful, considering firstly his 'Direct Argument' against names being ambiguous and secondly his criticism of Perry's argument against names being indexicals.

3.2.1 Pelczar's 'Direct Argument'

Pelczar's direct argument against names being ambiguous is not successful. Pelczar states that if proper names were ambiguous the fact that someone was unable to understand some occurrences of a name would not stand against their linguistic ability in the same way that someone who does not understanding all occurrences of an ambiguous term such as 'bill' does not stand against their linguistic abilities. Pelczar argues that this is not the case; if someone does not understand any occurrence of the proper name they are classed as being linguistically incompetent regarding that name. This argument is not successful for two reasons. Firstly, Pelczar is forgetting that it is impossible (in this theory) for someone to not understand a sentence containing a name given that they understand all of the other terms in the sentence. Even if the person can not locate the particular naming convention, they will recognise the term as a proper name and will give the corresponding reflexive meaning and hence have an understanding of the sentence. If Pelczar is concerned with the incremental meaning of a sentence when the referent is fixed, it seems that names are ambiguous in exactly the same way as other ambiguous terms. Consider the following sentence again;

(7) The tractor drove across the bill.

Upon hearing or reading this sentence an individual tracks the various meanings of 'bill' with which they are acquainted before settling on the one which is most likely given the meaning of the rest of the sentence and the context of utterance. If the individual has no knowledge of 'bill' meaning a strip of land they have two choices. They can either conjure up a scenario in which a tractor drives over a five pound note or a ducks mouth somehow gets trapped under the wheels of the tractor or they can choose to express confusion and make further enquiries regarding this particular occurrence of 'bill'. Similarly, consider the following sentence

(8) Bill is a talented actor.

Upon hearing or reading this sentence an individual tracks the various conventions for the name 'Bill' with which they are acquainted. It may be that, given the context of utterance, one particular Bill seems to be the obvious referent or it may be that there is more than one possible referent within the context. If the hearer knows of no person who is named 'Bill' and fits this description the hearer must express confusion over the occurrence of 'Bill' in this context and make further enquiries. Pelczar's arguments against the ambiguity of names is not successful; names are ambiguous in much the same way as other ambiguous expressions.

3.2.2 Pelczar's Criticism of Perry

Pelczar's main arguments against Perry centre around disparity among standard indexicals.

(a) Lack of Specificity: If Perry were to argue that proper names can not be indexicals because, unlike proper names, each indexical has a specific rule, Pelczar could argue back using 'this' and 'that' as examples of indexicals which seem to be indistinguishable therefore could be said to share a rule. However, this fails for two reasons. Firstly, this is not what Perry argues and secondly, even if this was Perry's argument it is not clear that 'this' and 'that' are indistinguishable; used appropriately 'this' refers to objects which are in some sense 'immediate' to the speaker whereas 'that' is intended to refer to objects which are apart from the speaker and the linguistic rules can reflect this fact.

(b) Messiness: If Perry were to argue that proper names cannot be indexicals as proper names are 'messy' and associated with complex social conventions, while indexicals have tidy rules to identify the referent, then Pelczar could argue back using an indexical such as 'here' which proves to be very messy when it comes to fixing the exact reference. Again, this is not what Perry states and secondly it can be argued that it is not the exact reference of 'here' which is unclear but the extent of the area surrounding the reference which must appeal to speakers intentions. It seems that any use of 'here' will always refer to the location of the utterance or inscription and this could be dictated by the linguistic rule. It is a further contextual role that determines the area around the location of the occurrence which is relevant to a particular utterance.

(c) Underdetermination: If Perry were to argue that the content of indexicals is always fully determined, while the content of a proper name is always underdetermined, Pelczar could argue back using an indexical such as 'that', the content of which is always underdetermined according to Pelczar. However, as Perry has distinguish between the reflexive and incremental content of indexicals it is unlikely that he would simply state that the content of indexicals is always fully determined.

Quite simply, these are not Perry's arguments, they are Pelczar's interpretation of what he takes to be the only argument that Perry has. That fact that Pelczar reinterprets what is already a clear argument of Perry's arouses suspicion. There are two matters of concern; firstly, do these interpretations fully capture Perry's main argument and secondly, are there other arguments within Perry's text which Pelczar has not addressed?

3.3 Perry's argument without reinterpretation

Perry states that the meaning of an expression of an indexical determines 'that certain specific contextual relationships to the utterance and utterer determine designation'. The primary point of Perry's argument is that the meaning of the indexical determines the contextual relationship whereas the meaning of a proper name does not determine the contextual relationship; this is determined by non-linguistic or extra-linguistic factors. Pelczar does not address this issue directly. The role that context plays in terms of indexicals is one regarding which Perry is quite clear. He does not elaborate in 'Reflexivity, Indexicality and Names' (the focus of Pelczar's argument) but in 'Indexicals and Demonstratives', published around the same time, Perry gives a detailed account of the role of context specific to indexicals, an account which excludes proper names. Perry distinguishes between pre-semantic and semantic context. We can consider Perry's example;

(9) I saw her duck under the table.

Knowledge of the context of utterance would help us to decide whether the person referred to had lost a pet, or was seeking refuge. In the 'duck' example it is the environment of utterance (pre-semantic) which determines what is said; this is not the case with indexicals, where the context is given (to a sufficient degree) by semantics. Regarding proper names, it is clear that they fall into the pre-semantic category; the environment of utterance helps to disambiguate, not the meaning of the term. It is this argument which Pelczar must address if he is to maintain that proper names are indexicals; he does not do this therefore Perry's distinction still stands.

3.4 Further Arguments Against Names Being Indexicals

Pelczar claims that the argument he considers is the only distinction that Perry makes between proper names and indexicals but this is not the case.^[43] There is another crucial difference between reflexivity and indexicality. Section 1.3 considered how proper names and indexicals are alike in that they can both be understood as having 'reflexive truth conditions' which only require that the hearer have an understanding of the linguistic meaning or rules attached to the term in order to be linguistically competent and to state conditions under which the proposition would be true. However, when considering the 'incremental content', that is the full meaning with the context fixed, the distinction between the two terms becomes very clear. Consider this example; the name 'John Perry' has been heard by a girl in a sentence and she understands the sentence even though she has no idea who the term 'John Perry' refers to, as she understands the meaning of proper names. Later that day the girl is introduced to John Perry, and she learns that he is the same John Perry that she had heard being referred to earlier. What happens now that she has all of the information and the full content of the earlier proposition is, the reflexivity disappears; the meaning is fixed. However, an expression containing an indexical expression retains its reflexivity even when the meaning is fixed.^[44] Once this distinction is made it is clear that, although names and indexicals have reflexivity in common, names are not indexicals.

3.5 Problems with Pelczar's Indexical Theory of Names

Although Pelczar's criticisms of the Reflexive Theory of Proper Names were not successful it could be the case that he provides independent motivation for choosing the Indexical Theory over the Reflexive Theory. Therefore, this section turns a critical eye on the positive aspects of Pelczar's theory.

3.5.1 Pelczar's 'dubbings-in-force'

In order to say that names are indexicals, Pelczar suggests that we add to our standard notion of contextual features {utterer, time, spatial position, possible world}, a fifth contextual feature; 'dubbings-in-force'^[45]. The 'dubbings-in-force' are just like Recanati's naming conventions or Perry's permissive conventions, but by making proper names indexicals, Pelczar needs to show that the contextual factors involved in determining the reference of proper names are no different from the contextual factors that we associate with standard indexicals, therefore 'dubbing-in-force' become part of the contextual factors. However, this move seems to work against Pelczar as it becomes immediately obvious that 'dubbings-in-force' are entirely different from the other contextual factors. The standard factors {utterer, time, spatial position, possible world} are always salient; they are objective, easily accessible and simply....there! This is not the case with 'dubbings-in-force'. It may be argued that as soon as a name is mentioned the relevant 'dubbings-in-force' become salient, but they are brought into the context by the use of the name whereas the other contextual factors are already there, in a sense waiting to be used by an indexical. The standard contextual factors are objective; if Pelczar was to reply that 'dubbings-in-force' were also objective, this would give these social conventions an odd kind of 'Platonic' presence. Furthermore, even if we grant Pelczar this fifth contextual factor, names could often still be ambiguous and he would still have to rely on the contextual role that Perry describes. Consider this example; two teachers are having a conversation and it just so happens that both of them know the same four people called 'Steven Jones'. Furthermore, all of these Steven Jones' happen to be relevant to this particular conversation, perhaps they are all pupils at the school who have done well in a recent exam. During the conversation it will be very difficult for the teachers to know which Steven Jones is being referred to as all four of the 'Steven Jones' dubbings-in-force will be salient. The name is undoubtedly ambiguous and dubbings-in-force being contextual factors are not going to make the name less ambiguous; the only way to disambiguate is to look at further features of the context ala Perry.

3.5.2 Motivation for the Indexical Theory

Despite the problems identified with Pelczar's theory above, if Pelczar could offer some strong motivation for choosing his Indexical Theory over the Reflexive Theory, it may be beneficial to try and save the Indexical Theory. However, the most frustrating aspect of Pelczar's paper is that he appears to offer no positive argument for choosing his Indexical Theory of Names over the Reflexive Theory. It is clear from Perry's paper that we can allow our understanding of this directly referential term to benefit from the reflexive/incremental distinction while retaining a type distinction between proper names and indexicals. Furthermore, the problems associated with standard direct reference theories can be handled adequately by the Reflexive Theory. So the question remains, what motivation is there for taking this extra step and saying that proper names are indexicals? Given this, and the fact that Pelczar has to force the extra contextual feature 'dubbings-in-force', Pelczar's argument that names should be included among the class of indexicals is less than convincing. Therefore, in conclusion of this final section, it is clear that Pelczar's attempt fails as firstly, he does not adequately address the differences between proper names and indexicals; secondly he attempts to equate social naming conventions with standard objective contextual factors and finally, he does not motivate his position by offering a positive reason for advocating his strong Indexical Theory of Proper Names over the Reflexive Theory of Proper Names. It has been shown that, although proper names can be said to resemble indexicals to the extent that they share an element of reflexivity, names are not indexicals.

Conclusion

The purpose of this paper was to consider to what extent proper names can be said to resemble indexicals. In Section 1 essential background for the paper was given in the form of brief introductions to direct reference theory for proper names, a theory for indexicals and a summary of an indexical theory of names. Section 2 focused more closely on three philosophers who compare proper names with indexicals but to different extents; of particular interest was a paper by Michael Pelczar who claims that proper names are indexicals. In Section 2 it became clear that the area of interest for this paper was not to criticise the 'indexical theory' as a whole, but to focus on the different extent to which the three philosophers were willing to compare proper names and indexicals. Section 2 concluded that, although Perry and Recanati's papers differ in attitude towards the comparison between proper names and indexicals, their theories are very similar and it was suggested that this shared theory be called the Reflexive Theory of Proper Names.

The final section of the paper directly addressed the question; do names resemble indexicals or is it feasible to go further and say that names are indexicals? The section focused on a debate between Perry and Pelczar. It became clear that Pelczar's attack of Perry's argument against names being indexicals is not successful. Firstly, Pelczar reinterprets Perry's argument and in doing so does not directly address the issue. Secondly, Pelczar does not address other arguments that Perry gives which show that names and indexicals are fundamentally different. It is concluded that Pelczar's criticism of Perry's argument has failed. However it may have been the case that, despite his unsuccessful attack on Perry, Pelczar's strong indexical theory is well motivated, therefore the positive aspects of Pelczar's paper are considered. Unfortunately, Pelczar proposes no benefits to be gained by choosing his Indexical Theory of Proper Names over the Reflexive Theory of Proper Names. Furthermore the inclusion of 'dubbings-in-force' as a fifth contextual factor, something which is essential to his strong theory, is less than convincing. Therefore, this paper concludes that although it may be beneficial to compare proper names with indexicals as proposed by the Reflexive Theory of Proper Names, the claim that proper names are indexicals has proved to be unsuccessful.

BIBLIOGRAPHY

Evans, G. (1973) 'The Causal Theory of Names' Aristotelian Society Supplementary 47

Evans, G. (1982) The Varieties of Reference, J. McDowell (ed.). Clarendon Press: Oxford.

Kaplan, D. (1989) 'Demonstratives: An Essay on the Semantics, Logic, Metaphysics and Epistemology of Demonstratives and Other Indexicals', in J. Almog et al. (eds), Themes from Kaplan. Oxford University Press: Oxford.

Kaplan, D. (1990) 'Words', Aristotelian Society Supplementary Vol. LXIV

Kripke, S. (1972) Naming and Necessity. Harvard University Press: Cambridge

Kripke, S. (1979) 'A Puzzle About Belief'. in N. Salmon and S. Soames (eds.), Propositions and Attitudes. Oxford University Press: Oxford.

Pelczar, M. (2001) 'Names as Tokens and Names as Tools'. Synthese 128

Pelczar, M. and Rainsbury, J. (1998) 'The Indexical Character of Names'. Synthese 114

Perry, J. (1997) 'Reflexivity, Indexicality and Names', in W. Kunne (ed.), Direct Reference, Indexicality and Propositional Attitudes. CSLI-Cambridge University Press: Stanford.

Perry, J. (1997b) 'Indexicals and Demonstratives', in R. Hale and C. Wright (eds), Companion to the Philosophy of Language. Blackwell: Oxford.

Perry, J. (1990) 'Individuals in Informational and Intentional Content', reprinted in Perry, The Problem of the Essential Indexical and Other Essays. Oxford University Press: New York

Recanati, F. (1993) Direct Reference: From Language to Thought. Blackwell: Oxford.

Taylor, K. (1997) 'Francois Recanati's Direct Reference: From Language to Thought Accomodationist Neo-Russellianism. Nous 31:4

^[7] Kripke gave many arguments against names being descriptive, one of which was the fact that descriptions are normally concerned with contingent facts e.g. 'The current leader of the Labour Party', whereas counterfactual statements require us to keep our finger on or 'rigidly designate' an individual, therefore the role of names is to rigidly designate; the referent must stay the same while the contingent facts can change. A rigid designator is a singular term which denotes the same thing in every possible world, where that thing exists. See Kripke (1972)

^[9] There are also indexical elements in words that indicate tense and other words such as 'local'.

^[11] There are various ways of formulating a rule for 'I'; this one come from Kaplan and works providing it is understand that different utterances may have different contents and that 'I' is in each of its utterances directly referential.

^[13] What is outlined is a brief summary of the Kaplan/ Perry theory of indexicals. Recanati has advanced this theory by including a third distinction. In addition to 'character' and 'content', Recanati offers a persuasive case for a 'psychological mode of presentation', a way of thinking about oneself (in the case of 'me') which is crucial in order for us to account for the difference in cognitive content between (1) 'An angry bear is about to attack me' and (2) 'An angry bear is about to attack Paula Milne'. (See Recanati, 1993, 135) Although interesting this distinction is not essential to this paper.

^[14] Kaplan provides only one level of content, 'official content'; this distinction between reflexive and incremental content (which is the same as 'official content') is Perry's.

^[15] This is Recanati's terminology, employed at this stage to refer to the three theories which are the focus of this paper.

^[16] 'Non-linguistic' means not part of the meaning of the term if the meaning of the term is concerned only with the linguistic type and not any token of that type.

^[19] This is Recanati's term; Perry calls it a permissive convention

^[20] As stated in the introduction, it is not the purpose of this paper to critically evaluate the indexical view of proper names as a whole; this would be an interesting topic but would require more discussion than this dissertation allows. This section is included as background as some of the problems discussed in this section are relevant to the history of direct reference theories.

^[31] Pelczar states that a 'dubbing' is a speech-act whereby a name acquires a referent, and a dubbing is 'in force' in a given context if in that context the item that was dubbed in that dubbing bears the name it received in that dubbing.

^[32] A criticism of this follows in Section 3.5.1

^[36] A detailed discussion of Pelczar's theory follows in Section 3

^[37] 'bill' is ambiguous between (i) a statement of money owed (ii) a law proposal (iii) a poster or leaflet (iv) a bird's beak (v) a narrow strip of land (vi) a nickname for the police!!

^[38] Criticism of this argument follows in section 3.2.1

^[43] Perry doesn't actually state 'this is another argument against the indexicality of proper names', but it seems clear that he considers it to be a distinction between the two types of terms, therefore Pelczar must consider it.