CEPA eprint 3849

Butter knives and screwdrivers: An intentionalist defense of radical constructivism

Alward P. (2014) Butter knives and screwdrivers: An intentionalist defense of radical constructivism. The Journal of Aesthetics and Art Criticism 72(3): 247–260. Available at http://cepa.info/3849
Table of Contents
Stecker’s dilemma
Encoding and decoding
Meaning declarations
Imputation
Conclusion
Robert Stecker has posed a dilemma for the constructivist theory of interpretation: either interpretations consist of statements with truth values or they do not. Stecker argues that either way, they cannot change the meaning of an artwork. In this article, I argue contra Stecker that if interpretations consist of meaning declarations rather than statements, they can change the meanings of the objects toward which they are directed, where whether they so consist is largely a function of the interpreter’s intentions. Hence, the second horn of Stecker’s dilemma is defeated.
Constructivism is the view that, rather than uncovering the meaning or representational properties of an artwork, an interpretation instead generates an artwork’s meaning. The more radical variants hold that the process of interpretation actually generates a new object – distinct from the products of the artist’s creative activity – which is what really gets interpreted. Robert Stecker argues, however, that the constructivist theory of interpretation is not tenable.[Note 1] In particular, he poses a dilemma for constructivism and claims that on neither horn can an interpretation change an artwork’s meaning.
The central goal of this article is to rescue constructivism – even in its more radical form – from Stecker’s dilemma. This is not, however, part of a more general defense of constructivism, understood as a theory of either what interpretation necessarily does or ought to consist in. Rather, it is part of a more general defense of meta-interpretive pluralism, the view that there are a number of distinct interpretive projects – including constructivist meaning generation – in which one might reasonably engage and which are neither incoherent nor illegitimate. Moreover, as will become apparent later, the defense of constructivism on offer here presupposes the legitimacy of intentionalist interpretative projects, including those which proceed by attempting to uncover the actual artist’s meaning intentions.
This article consists of four parts. First, I present Stecker’s dilemma and consider a number of responses that have been made to it. Second, I present a general account of encoding texts with meaning – including already meaningful texts. Third, I develop a novel response to Stecker’s dilemma according to which constructivist interpretations consist predominantly of meaning declarations rather than assertions. And finally, I consider the extent to which the kind of constructivism defended in the third part is superseded by the more traditional imputational version of the view.
Stecker’s dilemma
The central issue here is in what the interpretation of representational artworks consists. An interpretation gives a specification of a work’s meaning.[Note 2] At a general level, this includes a specification of both “truth in the work” and “truth through the work,” that is, what is going on in the world of the work and what claims, suggestions, recommendations, and the like are made about the actual world by means of the work. At a more particular level, this may include a specification of the meaningful elements of the work, what the meanings of these elements are, and the mode of meaning at issue, that is, whether the meaning should be taken literally, metaphorically, ironically, and so on. Examples of interpretive utterances include “American Gothic depicts a man and his daughter rather than a man and his wife,” “Gulliver’s Travels is a critique of policies designed to further scientific progress,” and “Withnail’s recitation of the lines ‘Man delights me not; no nor woman neither; no nor woman neither’ is an expression of his homosexuality.” Finally, although an interpretation can consist in an unexpressed psychological attitude toward a work, my focus here is on linguistically articulated specifications of meaning.
Stecker distinguishes between radical and moderate versions of constructivism. As Stecker characterizes it, moderate constructivism is the view that interpretations change the properties of artworks.[Note 3] Although Stecker does not make this entirely explicit, it is reasonably clear that the properties of artworks that interpretations are supposed to change are their representational properties. But Stecker is quite explicit that, according to moderate constructivism, interpretations effect a genuine or robust change in the representational properties of artworks themselves and not merely a trivial change in what meanings artworks are thought to have, which leaves the meanings they in fact have unaltered.[Note 4] It is worth noting that, so characterized, moderate constructivism can be understood in at least three different ways: first, an interpretation can be understood to change a previously meaningless work by giving it a meaning; second, an interpretation can be understood to involve replacing the meaning of previously meaningful work with a new meaning; third, an interpretation can be understood to involve adding an additional meaning to an already meaningful work. So, for example, an utterance of “American Gothic depicts a man and his daughter” might change a work depicting nothing into a depiction of a man and his daughter, it might change a depiction of a man and his wife into a depiction of a man and his daughter, or it might change a depiction of a man and his wife into both a depiction of a man and his wife and a depiction of a man and his daughter. Although Stecker gives little indication whether he prefers the second understanding over the third or vice versa, I suspect he would treat the first understanding as a variant of radical rather than moderate constructivism.
Radical constructivism, as Stecker casts it, is the view that rather than changing the properties of a preexisting artwork, interpretation creates a new art object. The picture here seems to have two separate elements. First, what artists produce are meaningless texts (broadly construed), and interpretations supply these texts with their meanings. And second, the meaningless texts artists produce are in a sense incomplete and become full artworks only when imbued with meanings by their interpretations. Hence, an interpretation of a meaningless text completes it and thereby turns it into an artwork by supplying it with a meaning.[Note 5] Stecker presents two central arguments against constructivism: one directed toward radical constructivism; the other directed toward moderate constructivism. It is worth emphasizing, however, that the latter argument poses equally serious difficulties for radical constructivism as it does for its more moderate cousin; it is presumably directed only toward moderate constructivism because Stecker no longer considers radical constructivism to be on the table at the point which it is offered. Stecker’s argument against radical constructivism is quite straightforward.[Note 6] Suppose that the radical constructivist is right that the products of artistic activity are meaningless texts.[Note 7] In lieu of some reason to believe that artistic activity is unique is this regard, it seems that we must also suppose that the products of critical activity – interpretations – are meaningless texts as well. And if interpretations are themselves meaningless texts, then they are ill suited to supply the products of artistic activity with meanings. In particular, in order to supply a text with a meaning, an interpretation minimally has to specify the meaning it is supplying the text with, but if the interpretation itself is a meaningless text, it does not specify anything at all.[Note 8]
Stecker’s argument against moderate constructivism takes the form of dilemma. The possibilities with which Stecker presents the constructivist are that interpretations make statements that are either true or false or that interpretations do not make statements that are true or false.[Note 9] And he argues that on neither possibility do interpretations change the meanings of the works they are interpretations of. Consider the first possibility: If an interpretation makes a statement attributing a meaning to a work that is true, then the work already had that meaning and, hence, did not acquire it as a result of the statement. If it makes a false meaning attribution, then the work lacks that meaning, and there is no reason to believe that a false attribution would subsequently cause the work to acquire it.[Note 10] Consider now the second possibility: If interpretations do not make statements that are true or false, then although they can change the reader or listener – or even the interpreter herself – in various ways, they cannot change the artwork itself. Suppose, for example, that rather than making a statement attributing a meaning to an artwork, an interpretation requests or invites or commands the reader or listener to believe or imagine the work has a certain meaning. Such an interpretation might well produce significant changes in the psychological states of the reader or listener. But it would not produce any genuine change in the artwork itself, except perhaps in the trivial sense in how the work is thought about.[Note 11]
It is worth emphasizing that the first horn of Stecker’s dilemma takes interpretations to have a pair of properties – making statements and being true or false – which are arguably independent. As a result, the second horn – which is, in effect, the negation of the first – includes three separate possibilities: interpretations make statements that are neither true nor false, interpretations consist of nonstatements which are neither true nor false, and interpretations consist of nonstatements which are true or false. Stecker does explicitly consider the first possibility and, in particular, Joseph Margolis’s view that interpretations make statements that are plausible or implausible, apt or inapt, or reasonable or unreasonable, rather than true or false.[Note 12] And he argues that even so understood, interpretations cannot change the meanings of artworks.[Note 13] But he does not seem to consider the third possibility of nonstatements that are true or false. An example might involve saying, “I bet ten dollars on the Habs” in response to someone’s prior commitment to accept any such bets: what you say is strictly true even though what you are doing is betting rather than stating.
It is also worth noting that, at least on the received view, the illocutionary point of a statement or an assertion is to induce belief in the expressed proposition or, more modestly, to commit oneself to its truth.[Note 14] As a result, the very same argument
against interpretations changing the meanings of their objects used by Stecker in the second horn of the dilemma could equally well have been deployed in the first: a statement which induces beliefs about the meaning of an artwork in a reader or listener no more changes the artwork than does a request or an invitation or a command that the reader or listener believe or imagine the work has a certain meaning.[Note 15] : 51–60, at p. 52). The upshot, in my view, is that the issue of whether or not interpretations have truth values is a bit of a red herring. The real issue is whether or not there is any kind of speech act which can be performed via an interpretation that can change the meaning of the work it is an interpretation of.
One might, of course, worry whether or not utterances designed to alter rather than uncover the meanings of the entities toward which they are directed even count as interpretations at all. After all, interpreting someone or something arguably consists in trying to understand or make sense of that person or thing, which, again arguably, one can do successfully or unsuccessfully. And acts of meaning construction fail on both counts: one cannot successfully understand something if one is not attempting to understand it in the first place. There are, however, two responses worth making to this worry here. First, strictly speaking, this article is a response to Stecker’s dilemma for constructivism. For the purposes of this argument, at least, Stecker concedes that were meaning construction of the kind discussed above possible, it would count as interpretation. His goal is to establish that it is not possible. As a result, a response to Stecker’s dilemma on behalf of constructivism does not need to establish that meaning construction counts as genuine interpretation, but only that it is possible. Second, since focusing narrowly on a response to Stecker would limit the interest of this article, it is worth saying a few things by way of addressing this worry head-on. Although the objects of interpretation are works of art and literature, the objects of a theory of interpretation are works of art criticism and literary criticism. And unless one is developing a revisionary theory of criticism – articulating an account of what, by some measure, interpretation ought to consist in rather than what it in fact does – one needs to be sensitive to actual practice. Hence, to the extent that critical practice includes critical works designed to construct meaning – perhaps alongside those designed to recover it – one will be hard pressed to deny that the former counts as at least a kind of interpretation. One might, of course, disapprove of certain interpretive enterprises, believing, for example, that real or genuine interpretation consists only in the recovering of meaning. But such disapproval no more shows that critical works aimed at meaning construction do not count as interpretations than one’s disapproval of certain trends in twentieth-century art shows that readymades such as Snow Shovel or Bottle Rack are not works of art.
There have been a number of responses to Stecker’s dilemma. Of particular interest in this article are those of Michael Krausz and Philip Percival.[Note 16] Krausz’s defense of the imputational version of radical constructivism is the focus of Section IV. In the remainder of this section, Percival’s response will be addressed. Percival’s strategy is to attack the first horn of Stecker’s dilemma – that if an interpretation makes statements that are true or false, it cannot change the meaning of its target work.[Note 17] In particular, Percival argues that what he calls “weak anti-individualist” moderate constructivism avoids the first horn of the dilemma. The view is characterized by three distinct tenets: first, contra “mainstream constructivism,” interpretations have truth values; second, only some, but not all, interpretations change their objects; third, an individual interpretation cannot change an artwork, but rather only a plurality of interpretations can jointly do so.[Note 18] Percival’s basic picture seems to be that what might be called the “original meaning” of an artwork is the product of the artist’s intentions, corresponding roughly to what she means by it. Any individual interpretation which deviates from the artist’s intentions will initially be false, but a plurality of false judgments attributing the same meaning can render subsequent judgments to that effect true.[Note 19] The model here is ‘Madagascar.’ Originally a name of part of the African mainland, it now refers to an island off the east coast of the continent. The original uses of the name – or its historical predecessor – to refer to the island were erroneous; only when a practice of using it to refer to the island emerged did it undergo a semantic change and come to refer to Madagascar.[Note 20] More precisely, sometimes how a speaker intends to use a word can diverge from the word’s conventional meaning despite his sincere intention to use it in its conventional sense; if this divergent use becomes sufficiently embedded in the original speaker’s linguistic community, the word may undergo a change in meaning. Similarly, how a critic understands and “uses” an artwork can diverge from the work’s objective interpretation-independent meaning even when the critic intends to grasp and utilize the work’s objective meaning. And, as with language, if this divergent meaning becomes embedded in the critical community, the work may undergo a change in meaning.[Note 21] So, for example, if the view that Hamlet delays his revenge due to depression over the death of his father becomes sufficiently widespread among critics, then this view will replace Shakespeare’s original intentions on the matter as part of the meaning of Hamlet.
There are a couple of comments worth making about Percival’s argument. First, it remains unclear whether Percival is correct that if an erroneous interpretation of an artwork becomes sufficiently embedded in the critical community, the meaning of the work could undergo a corresponding change as a result or, following Stecker, this would simply count as a “trivial” change in the way the work is thought of.[Note 22] And the analogy Percival draws between language and art is of little help here. In the former case, meaning change occurs when a divergence between conventional sentence meaning and speaker meaning – or, perhaps better, utterance meaning and utterer’s meaning – becomes embedded in the linguistic community. But not only is the interpretation-independent meaning of an artwork more naturally modeled on speaker meaning – what the artist meant by her work – than any sort of conventional meaning, it is not clear in what sense a critic uses or means something by the object of her interpretation. Second, it is reasonably clear that Percival’s weak antiindividualist moderate constructivism is actually a version of interpretive realism rather than any kind of constructivism.[Note 23] On Percival’s picture, artworks have interpretation-independent meanings. Contra Stecker, these meanings are not a function of artists’ historical meaning intentions but rather of critics’ evolving meaning beliefs and, hence, can undergo change over time. Nevertheless, on Percival’s picture, the goal of interpretive practice is to uncover a work’s objective meaning (at the time), and individual interpretations can succeed or fail in this regard.
In what follows, I follow a rather different line of response to Stecker’s dilemma. First, contra Percival, I assume that, on the constructivist picture I defend, interpretations consist of nonstatements which may (or may not) have truth values. Hence, my target is the second horn of Stecker’s dilemma rather than the first. Second, again contra Percival, I defend an individualist account of constructivism according to which an individual interpretation can change an artwork – or, perhaps, even create a new art object. And third, although I agree with Percival that not every interpretation can change the meaning of its object, I am inclined to think that a broader range of interpretations can do so than Percival would likely accept.[Note 24] As long as relatively minimal formal constraints are satisfied – such as coherence, intelligibility, and the like – an interpretation can change an artwork, even if it is implausible, unreasonable, inapt, or “just crazy.”[Note 25] Unlike the view Percival defends, the view developed below uncontroversially counts as a species of constructivism.
Encoding and decoding
The guiding idea here is that artistic activity, as well as constructivist interpretation of the products of artistic activity, can be viewed as a kind of code making. In both cases, the upshot is an object encoded with meaning, which is largely a function of what the encoder meant by it and which is, at least in principle, capable of being uncovered by means of a realist decoding interpretation of it. The difference between an artist and a constructivist interpreter is that the former is a primary encoder, in the sense that she creates the object which she encodes with meaning, whereas the latter is a secondary encoder – encoding with meaning an object which not only did he not create but which has already been encoded with meaning by its creator. It is worth noting the analogy between artistic activity and code making should not be overstated: not all artists mean anything at all by the products of their artistic activity; even those who do have meaning intentions need not have the kind of communicative intentions that are characteristic of code making; artists’ meaning intentions are often subservient to their broader artistic and aesthetic goals.
The central question of this section is whether or not there is any reason to suppose, in a case in which someone encodes with meaning an object already encoded by its creator, that the meaning intentions of the primary encoder affect the representation properties of the text or the object out of which it is generated, but the meaning intentions of the secondary encoder do not.[Note 26] Consider by way of analogy an object created with the intention that it be a butter knife, that is, that it have the function of spreading butter. And suppose that after acquiring it, I intend for it to be a screwdriver, that is, that it have the function of tightening and loosening screws, as I have sometimes been wont to do. Now suppose that the intentions of the manufacturer suffice to make the object a butter knife. The question is whether there are any grounds to suppose my subsequent intentions fail to make it a screwdriver. There are a number of strategies one might deploy in order to establish that the manufacturer’s intentions suffice to confer a function on the object but mine do not. First, one might argue that one can confer a function on an object by means of one’s function intentions only if the object was previously functionless, but since the object at issue already had the function of spreading butter, my intention that it have the function of tightening and loosening screws could not make it so. The trouble with this is that people recycle old artifacts and give them new functions all the time. Someone might, for example, turn an old wooden spoon into a puppet. Second, it might be rejoined that such cases involve the modification of the previous objects. Hence, one might argue that one can confer a function on a object by means of one’s function intentions only if the object is modified in order to serve the function, but since I did not modify the butter knife, my intention that it be a screwdriver does not suffice to make it so. The trouble with this is that people again confer functions upon objects without modifying them all the time. Consider, for example, children who choose sticks they find on the ground with the intention that they be props in games of make-believe of various kinds, such as sword-fighting games. Finally, one might combine the first two strategies and argue that one can confer a function on an object one does not modify by means of one’s function intentions only if it did not previously have a function; since I did not modify my butter knife and it did have a prior function, my function intentions do not suffice. But if this were correct, it would render the unmodified found art – where a change of venue does not count as a modification – impossible.
Someone might object at this point that although a butter knife can be used as a screwdriver, it cannot literally be a screwdriver, that is, have the function of tightening and loosening screws. The reason is that the features that enable the butter knife to be successfully used as a screwdriver – such as having an edge that can fit into screw heads – are distinct and independent of the features that make it a butter knife – such as flatness. And since to have a function is to be designed to do something, the merely fortuitous features that enable a butter knife to be used as a screwdriver do not suffice to give the butter knife the function of tightening and loosening screws, regardless of anyone’s intentions, although I am not convinced that the fact that it is the undesigned accidental features of a butter knife that enable it to be used as a screwdriver shows that the butter knife cannot acquire the function of tightening and loosening screws. The point is moot here. My view is that it is the object serving as the butter knife that becomes a screwdriver and not the butter knife itself. In effect, the view on the table is that human artifacts, like butter knives and screwdrivers, are object–function pairs. As a result, even if the fact that it is the undesigned features of a butter knife that facilitate its use as a screwdriver shows that the butter knife cannot have the function of tightening and loosening screws, it does nothing to show that the object serving as a butter knife cannot have this function. Moreover, this maneuver does nothing to undermine the parallel between creating butter knives and primary encoding. After all, a representational artwork can similarly be viewed, not as an object–function pair, but as an object–meaning pair. And just as the primary encoder produces the particular text she does in order to generate something capable of meaning what she intends, the butter knife manufacturer produces an object with the pattern of features she does in order to generate something capable of serving the function she intends. And just as a secondary encoder can have different meaning intentions regarding the text produced by a primary encoder, someone can have different function intentions regarding an object originally created by someone else.
What I suggest is that in lieu of any special reason to think that meanings are different from functions in this regard, then the analogous conditional holds in the case of meanings, namely, if the meaning intentions of the primary encoder affect the representational properties of the text or the object out of which it is generated, then the meaning intentions of the secondary encoder do as well.[Note 27] There does not seem to be anything special about the relation between meanings and meaning intentions which could establish that, unlike function intentions, only meaning intentions regarding modified texts – or unmodified objects no one had previously meant anything by – could imbue those texts with meaning. One might worry that this allows any old meaning intention on the part of a secondary encoder to affect the meaning of the text, no matter how ill suited it might be to the text. There are two responses worth making to this worry. First, an original artist’s meaning intentions can be just as ill suited to the text he produces as can those of a secondary encoder. Moreover, at least in principle, the meaning intentions of a secondary encoder might be an improvement over the original. And second, my inclination is to allow a rather broad range of meaning intentions to count as meaning conferring, but to differentiate between those that are better and less well suited to the text: just as there can be better or worse screwdrivers, there can be better or worse codes. Of course, it is possible, despite one’s intentions, to fail to make a screwdriver at all, that is, to produce a new object, or to have the requisite functional intentions toward an existing object, not even minimally capable of fulfilling the function of tightening and loosening screws. Similarly, one can fail to make a code, despite one’s intentions to do so, if one’s meaning intentions render the text toward which they are directed unintelligible.
Meaning declarations
Recall that on the second horn of his dilemma, Stecker argues that if an interpretation does not consist of statements that are true or false, then it cannot change the artwork it is directed toward. As above, the central issue is not whether the sentences which make up an interpretation have truth values but rather whether or not the speech acts performed by means of them are capable of changing the representational properties of their objects. And Stecker’s view is that while nonstatements can bring about a robust change in the speaker or writer or her audience, they cannot do the same vis-á-vis an artwork toward which they are directed: “I cannot change an artwork by issuing a command, or by imagining something with regard to it.”[Note 28]
The basic difficulty with Stecker’s argument is that it is inadequately sensitive to the range of possible illocutionary actions that can be performed by means of an utterance. Following Searle, we might distinguish between five different types of illocutionary actions – representatives, directives, commissives, expressives, and declarations – the first four of which, arguably at least, cannot change an inanimate object toward which they are directed.[Note 29] Representatives are illocutionary acts designed to commit a speaker to the truth of the proposition expressed by the sentence he utters or, more strongly, to induce belief in this proposition in his audience. A statement is, of course, a kind of representative.[Note 30] Directives are illocutionary acts whose point is to get the audience to do something. These include invitations, requests, and commands, among other things.[Note 31] Commissives are illocutionary acts designed to commit the speaker to some future course of action and include most prominently promises.[Note 32] Finally, expressives are speech acts designed to express a speaker’s psychological states, such as thanking or congratulating.[Note 33] And Stecker is probably right that a speaker is no more likely to effect a robust change in an artwork – at least not a change in its representational properties – by simply requesting that her audience do something involving it, committing herself to do something involving it, or expressing her attitudes toward it than she is by committing herself to the truth of a proposition concerning it.[Note 34]
Nevertheless, the remaining type of illocutionary actions – declarations – can effect a change in inanimate objects toward which they are directed. The point of a declaration is to bring about a “correspondence” between the proposition expressed by a speaker’s utterance and the world, and if it is performed successfully, the speaker does thereby bring about this correspondence. As Searle puts it, “[declarations] bring about some alternation of the status or condition of the referred to object or objects solely in virtue of the fact that the declaration has been successfully performed.”[Note 35] Suppose, for example, an employer says, “You are fired,” to one of her employees. Now she might, of course, be thereby making an assertion; if so, then to perform her speech act successfully is minimally to commit herself to the truth of the proposition that the employee in question is no longer in her employ. And not only would she not thereby fire her employee, but her statement to that effect could even be incorrect. However, if she performs a declaration by means of her utterance, and her declaration is successful – she has the relevant authority and the various legal and contractual requirements governing the termination of employment are satisfied – then she thereby fires the employee, effecting a robust change in the latter’s properties.
The central idea here is that the specifications of a work’s meaning which make up an interpretation of it can be used to make meaning declarations and not only used to make statements. And if they are successfully so used, the interpretation thereby changes the representational properties of the work, encoding it (or its constituent text or the object from which the text was generated) with meaning. Suppose, for example, a critic says, “Withnail’s recitation of the lines ‘Man delights me not; no nor woman neither; no nor woman neither’ is an expression of his homosexuality.” Now, of course, the critic might be thereby making a statement and, if she is, the truth value of her utterance depends on whether or not it corresponds to the prior meaning of the relevant part of the film Withnail and I, which perhaps is a function of Bruce Robinson’s intentions regarding the scene in question. But if she thereby makes a declaration and her declaration is successful, her speech act changes the representational properties of the film – or, perhaps better, the filmic text – making it part of the meaning of the film that Withnail’s recitation of the lines in question are an expression of his homosexuality regardless of Robinson’s own intentions. As above, one can encode a meaningful work – one whose original author meant something by – with a new meaning; hence, the successful performance of a meaning declaration does not require that the object of some such declaration be previously meaningless. Finally, it is worth reiterating that if I am right that critics can make meaning declarations by means of the specifications of meaning that constitute their interpretations, then Stecker is wrong that interpretations cannot change the representational properties of their objects, and, hence, the second horn of his dilemma is undermined.[Note 36]
In general, in order to make a declaration by means of an utterance, a speaker needs to occupy the relevant role. So, for example, just as one needs to be an employer to terminate someone’s employment by means of saying, “You are fired,” one needs to be an umpire in order to end someone’s turn as a batter or runner by means of saying, “You are out,” and one needs to be a judge to make someone criminally liable for his actions by means of saying, “You are guilty.” Now, of course, occupying the relevant role is by itself not sufficient for making a declaration: after all, an occupant of such a role remains capable of making a statement by means of her utterance. What is minimally required in addition is that the speaker have the requisite intentions – the intention to fire someone or end his turn or make him criminally liable as opposed to intention to commit oneself to the truth of the corresponding propositions. But in some cases, declarations can be made without having to occupy any given role. For example, anyone can declare something to be a prop in a game of make-believe.[Note 37] If I say of a stick “This is a sword” with the requisite intentions, I thereby make it a prop that generates sword imaginings in my game of make-believe without my needing to occupy any sort of authoritative role. And encoding is like “prop making” in this respect: one does not need to occupy any particular role to encode a text with meaning. Anyone whatsoever can encode a text with meaning by means of making meaning declarations, even texts that they did not produce. In order for a couple of kids to make a code, for example, they do not need any special authority. In fact, one does not even need an audience: one can encode a text with meaning as a private joke for one’s own amusement. All that is required is that one’s utterances be made with the right kinds of intentions. In effect, what we are left with is a kind of meta-interpretive intentionalism. According to the more familiar object-level interpretive intentionalism, the correct interpretation of an artwork is a function of the artist’s intentions. According to meta-interpretive intentionalism, in contrast, whether an interpretation consists of statements designed to decode a work’s preexisting meaning or declarations designed to encode a work or its constituent text with meaning is a function of the critic’s intentions: if he intends to commit himself to the truth of the propositions expressed by the sentences he utters, then he makes statements by means of his utterances; if he intends to thereby encode the work or text with meaning, he makes meaning declarations by means of his utterances.
Now it is worth noting that in some cases, one can have both the relevant authority and intentions to make a declaration without successfully doing so.[Note 38] Someone might, for example, occupy the employer role and intend to terminate someone’s employment by means of saying, “You are fired,” and yet still fail to fire the employee in question because certain legal or contractual requirements governing terminations have not been met. In the case of meaning declarations as well, having the requisite intentions is not by itself sufficient. But what is required in addition for an attempted meaning declaration to be successful is a relatively minimal intelligibility requirement. What this requires is that there be a principled way of subsequently decoding the putatively encoded text – perhaps by means of some kind of translation algorithm – and that the encoded meaning at least not be overtly contradictory.
At this point, however, someone might raise the following worry: if interpretive utterances are declarations rather than statements, it is unclear whether and how they can be genuine bearers of truth. After all, it is arguably nonsensical to attribute truth values to at least some nonstatements; expressives, such as “ouch,” are a particularly clear example. The first thing to note is that insofar as this article is understood narrowly as a response to Stecker, the question is moot. After all, Stecker concedes, at least for the sake of argument, that interpretive utterances need not have truth values. But again, if this article is to be of broader interest, this worry does need to be addressed head-on. Consider, again, an utterance of “You are fired.” Now if the speaker thereby makes a statement, then it is clearly capable of truth or falsity: if, at the time, the subject of the utterance has lost his job then it is true; if, at that time, the subject retains his position, it is false. But suppose instead that the speaker thereby performs a declaration. The question is whether or not under such circumstances the utterance is a candidate for truth or falsity. The first thing to note is that although the illocutionary point of the utterance is different when it is used to make a declaration than when used to make a statement, the propositional content remains the same: both utterances express a proposition which is true only if the subject has lost his job. One might worry that in the case of declarations these truth conditions are both incidental to the point of the utterance and trivially satisfied, but in my view this is mistaken. The point of the utterance is, in fact, to make the proposition expressed by it true. Moreover, this proposition is not trivially true. The subject of a declarative use of “You are fired” might, after all, reasonably reply by saying, “That’s not true” or “No, I am not.” And what he would presumably be claiming is that the legal or contractual requirements for successful termination have not been met, or that the speaker lacks the relevant authority. It is worth noting, however, that the proposition expressed by a declarative use of “You are fired” must be understood to contain a temporal index – that the subject of the utterance lost his job as of the time of the utterance. After all, prior to the successful declaration, any assertive use of “You are fired” would have been false.
Consider now a declarative use of an interpretive sentence, such as “Gulliver’s Travels is a critique of policies designed to further scientific progress.”[Note 39] As with the prior case, this utterance expresses a proposition that is also expressed by at least some stating uses of the same sentence – a proposition that is true only if it is true of Gulliver’s Travels or its constituent text that it critiques the policies in question. But again exactly which proposition is expressed is context relative.[Note 40] Moreover, as with the previous case, the point of this meaning declaration is to make the proposition it expresses true. Unlike the firing case, however, the truth of this proposition is a relatively trivial matter exactly because the conditions that have to be met in order for a meaning declaration to be successful are so minimal. To a stating use of “Gulliver’s Travels is a critique of policies designed to further scientific progress,” the rejoinder “No, it is not” might reflect a substantial disagreement over Swift’s intentions regarding his novel, but to a declarative use of this sentence, the rejoinder can only consist in the charge that the intelligibility conditions governing acts of encoding have not been satisfied.
It is worth noting that, on the view on offer here, distinct critics’ meaning declarations regarding a single work or text can contradict one another as well as the meaning intentions of the original author. One critic might declare that a work, or one of its elements, means p, while another critic might declare or the original author might intend that it means not-p.[Note 41] This possibility poses potential difficulties for realist decoding interpretations of a work – those consisting of statements rather than declarations – made subsequent to the occurrence of the conflicting meaning declarations. After all, if one critic’s meaning declarations or the original author’s intentions make “x means p” true and a second critic’s declarations make “x means not-p” true, then, assuming that the latter is equivalent to the negation of the former, x means not-p iff not-(x means p), this leads to a straightforward contradiction.
One might argue that rather than supplementing meaningful works with additional meanings, constructivist interpretations replace the previous meanings of works with new meanings; hence, conflicting meaning declarations do not yield contradictory time-relative meaning claims. But not only does this render a work’s representational properties overly fragile – at constant risk of shattering in the hands of critics – it offers no help with the possibility of simultaneous conflicting meaning declarations. Alternatively, following Margolis, one might argue that interpretive claims of the kind at issue are apt, reasonable, or plausible rather than true.[Note 42] After all, the reasonableness, and so on, of both “x means p” and “x means notp” does not lead to any kind of contradiction even if their truth does do so. The trouble here is that given the intentionalist assumption being made here, the statements at issue are unquestionably true, whether or not they are also reasonable, plausible, or apt. Finally, one could argue that meaning claims are true only relative to interpretations and not absolutely so, and note that the fact that “x means p” is true relative to one interpretation does not contradict the fact that “x means not-p” relative to another. Although I take this to be the most promising strategy, a certain amount of delicacy is still required. In particular, an account of the sense in which the truth of a meaning claim is relative to an interpretation needs be given on which this does not simply amount to its being true according to this interpretation. After all, the latter would sever a work from its meaning in a way which many would find quite unsatisfying. Consider, by way of analogy, the difference between “Mary is tall relative to Fred” and “Mary is tall according to Fred.”
One way to sidestep these difficulties altogether would be to abandon the moderate constructivist picture according to which a constructivist interpretation changes the representational properties of an enduring work in favor of the radical constructivist picture according to which an interpretation creates a new work. The basic idea is that a work is, roughly, a text–meaning pair, and insofar as a single text is paired with distinct meanings – via an original artist’s meaning intentions and subsequent critics’ meaning declarations – it is a constituent of multiple distinct works. As a result, the object of a decoding interpretation is not one work with a plurality of meanings, but rather one of a plurality of works, each of which shares a common text and which has a univocal (intended) meaning. The upshot is that rather than grounding the truth of incompatible meaning claims regarding a single work – “x means p” and “x means not-p” – on the radical constructivist picture, incompatible meaning declarations ground the truth of compatible meaning claims regarding distinct works – “x means p” and “y means not-p.”
One might worry, however, that this maneuver falls prey to Stecker’s first argument, which was directed specifically against the radical variant of constructivism. Recall that Stecker argued that any reason to suppose that the products of artistic activity are meaningless texts – which need to be imbued with meaning in order to become completed works – is also a reason to suppose interpretations are themselves meaningless texts. But if interpretations are themselves meaningless, they are incapable of imbuing the products of artistic activity with meaning. On the picture on the table, however, the products of artistic activity are not meaningless texts but rather completed works imbued with meaning by artists’ meaning intentions. Similarly, interpretations consist of sentences imbued with meaning by critics’ illocutionary intentions. And rather than turning meaningless texts into works, what constructivist interpretations do is create new works which share texts with the works artists create.
Imputation
The constructivist picture on offer here presupposes that there are objects of interpretation which are both “complete and independent of interpreters,” and which can be subjected to either encoding or decoding interpretations.[Note 43] It has been argued, however, that objects of interpretation are interpretation-dependent entities – “text/interpreter complexes” – which emerge when the interpreter imputes salience, and perhaps other properties, to elements of the text.[Note 44] Moreover, on this view, the text itself lacks any kind of complete and determinate nature independent of the text/interpreter complexes of which it is a constituent: how it manifests itself to interpreters is itself a product of the interpretive process.[Note 45] If this is right, then the construction of an object of interpretation occurs prior to any meaning declarations that might be made vis-a’ -vis some such object, rendering the encoding process I have been defending a kind of secondary constructivist afterthought.
Stecker objects to the imputational picture on the grounds that it (putatively) entails that every interpretation generates a distinct object of interpretation – at least those that impute salience to different features of a text – and, hence, that there cannot be competing interpretations of a single work.[Note 46] After all, if objects of interpretation are text/interpreter complexes, then interpretations offered by distinct critics inevitably have distinct objects. Krausz’s response to this worry is to suggest that objects of interpretation generated by distinct interpretations – and imputing salience to distinct elements of a text – can nevertheless stand in a relation of numerical identity to one another. Just as diachronic identity is compatible with a difference in properties over time, Krausz argues that identity of objects of interpretation is compatible with a difference in properties “across” interpretations.[Note 47]
Stecker balks at this suggestion because he takes it to violate the following condition on strict identity:
a = b iff, for all times t, all and only properties of a at t are properties of b at t.48
After all, unless objects of interpretation are fleeting existents on the imputational model, distinct interpretations imputing different properties to a text will generate objects of interpretation having different properties at a given time. Krausz’s rejoinder is to reject the idea that identity must be strict or “absolute” on the grounds that it is incompatible with change over time.[Note 49] In my view, Krausz is simply wrong on this point: as Stecker formulates it, strict identity only requires exact sameness of properties at any given time, not at distinct times. Nevertheless, it remains open to Krausz to simply reject Stecker’s formulation of strict identity. In particular, Krausz could reasonably argue that, just as the requirement of sameness of properties is relative to both time and location in the case of four-dimensional physical objects, the requirement of sameness of properties is relative to both time and interpretation in the case of objects of interpretation. That is, rather than Stecker’s time-relative condition on strict identity, Krausz might offer the following time/interpretation-relative condition:
a = b iff, for all times t and interpretations i, all and only properties of a at t relative to i are properties of b at t relative to i.
And if this is right, the fact that objects generated by distinct interpretations can differ in properties at a given time does not entail that they are numerically distinct objects of interpretation.
Of course, it is one thing to be told that the identity of objects of interpretation is compatible with interpretation-relative differences in properties, but it is another to be told exactly what conditions have to be satisfied in order for objects generated by distinct interpretations to be numerically identical to one another. The obvious suggestion is that what is required is that there be a shared “further object” – a text or, borrowing from Paul Thom, a “configuration” – to certain of whose elements or features both interpretations impute various saliencies.[Note 50] Even if this further object is, in some sense, neither complete nor interpretation independent, it can nevertheless serve as a ground for identities among text/interpreter complexes which share it. Krausz, however, rejects the view that some such interpretation-independent object can serve this role. In order to do so, it must be the same text or configuration that is shared by the objects of interpretation at issue. But, Krausz argues, such further objects themselves lack conditions of individuation, and, hence, there is no fact of the matter, in any given case, whether or not the same text/configuration is shared by the objects of interpretation at issue.[Note 51]
Before addressing Krausz’s argument here, it may prove fruitful to make explicit a number of distinctions that have been heretofore implicit in this discussion. The first distinction is between what might be called “products of artistic activity” and texts. The products of artistic activity are the entities – the objects or events – artists produce, considered independently of anything they or anyone else might mean by them. So, for example, the product of a painter’s artistic activity might be a canvas with various painted lines and colors on it, and the product of an author’s artistic activity might consist of a large number of marks on numerous sheets of paper. As I understand it, this is what Thom means by a configuration. A text, in contrast, is a pattern of meaningful elements, such as a sequence of words and sentences. A text can be generated from a product of artistic activity if the artist, or someone else, means something by some of the elements that make it up. Again, as I understand it, this is more or less the role played by the imputations of salience that Krausz invokes. A second distinction is between texts and works. Works are interpreted texts or text–meaning pairs – patterns of meaningful elements together with the meanings of both the elements that make them up and the texts as a whole. As such, despite having texts as constituents, works are nevertheless distinct from them. Moreover, on this picture it is at least in principle possible for a single text to be shared by distinct works. A final distinction is between works and objects of interpretation. Although a work may be an object of interpretation, exactly what the object of interpretation is in any given case depends on the nature of the interpretive project. Insofar as the interpretation consists in decoding, the object of interpretation has to be something which has a meaning to be uncovered; hence, the object of a decoding interpretation can only be a work, and not the text which in part constitutes it or the product of artistic activity from which the text was generated. But if the interpretive project consists of constructivist meaning generation, the object of interpretation could be (i) the work, if constructivist interpretation merely changes the meaning of its object, as the moderate constructivist would have it; (ii) the text, if constructivist interpretation transforms its object into a work; or (iii) the product of artistic activity, if constructivist interpretation generates both a work and the text which in part constitutes this work.
Now the trouble with Krausz’s argument is that it equates interpretation-independent entities with unconceptualized reality – reality in itself unrepresented by spatio-temporally situated thinking and perceiving subjects. Even if works and texts are interpretation-dependent entities, the products of artistic activity are not.[Note 52] But if they are unconceptualized in this sense, then Krausz is probably right that they cannot be individuated; the notion of individuating aspects of unconceptualized reality may even be incoherent. But to say that an entity is interpretation independent does not entail that it is concept independent; that is, even if the existence and nature of an entity does not depend on being the object of an interpretation, it might so depend on satisfying certain concepts – concepts grounded in the shared linguistic practices of a human population or, perhaps, even in the evolutionary history of the human species. In particular, one might argue that to be a product of artistic activity is to be correctly conceived of as an entity which was created by an artist, consisting of a pattern or configuration of elements – some of which may be conventionally (or symbolically) meaningful – whose maker may have meant something by it.[Note 53] And not only do such concept-dependent entities have relatively clear conditions of individuation, conceiving of an entity in this way simply does not require interpreting it in any nontrivial sense. This, of course, clears the way for products of artistic activity, so understood, to ground the identity relation between imputed objects of interpretation. But, more importantly, it obviates the need for an identity relation between such entities: insofar as they are interpretation-independent but concept-dependent entities, there is no impediment to taking the products of artistic activity themselves to serve as objects of constructivist interpretation.
Conclusion
The general goal of this article is to motivate metainterpretative pluralism, the view that there are a number of distinct interpretive projects in which one might reasonably engage which are neither incoherent nor illegitimate. The products of artistic activity are the common objects of these various enterprises – to the extent that they have a common object – and are meaningful in a number of different senses: they often contain elements which are conventionally or symbolically meaningful; the artists who create them and the critics who interpret them mean various things by them and their elements; they have intention-independent meanings justifiably attributed to them on the basis of some notion of fit; and so on.[Note 54] And none of these counts as the unique meaning of a work. As a result, having any one of these as the target of a decoding interpretation of a work is perfectly legitimate, and which of these a critic pursues is simply a matter of his interpretive interests.
The particular goal here is to defend an interpretive project wherein a critic disregards the various prior senses in which a work is meaningful and generates her own meaning for it via a series of meaning declarations.[Note 55] Like the various kinds of decoding interpretations mentioned above, this enterprise is both coherent and legitimate, and whether an interpreter engages in it is a matter of his interpretive interests. My ultimate claim here is not that critics commonly do engage in interpretive encoding but only that they could do so if they so desired.
Endnotes
1
Robert Stecker, “The Constructivist’s Dilemma,” The Journal of Aesthetics and Art Criticism 55 (1997): 43–51.
2
The word ‘specification’ is intended to be neutral between various accounts of interpretation.
3
Stecker, “The Constructivist’s Dilemma,” p. 43.
4
Stecker, “The Constructivist’s Dilemma,” p. 44.
5
One might argue that although a representational artwork consists of a text – here understood as a configuration of meaningful elements – and a meaning, the text which in part constitutes the work is itself generated by interpretation. If this is right, then the products of artistic activity are not texts but rather entities which are turned into texts by interpretations. This issue is addressed in Section IV below.
6
Stecker, “The Constructivist’s Dilemma,” pp. 43–44.
7
Or, again, entities turned into texts by interpretations.
8
One might also worry whether meaningfulness is required for an interpretation to be of a particular text in the first place.
9
Stecker, “The Constructivist’s Dilemma,” p. 44.
10
Stecker, “The Constructivist’s Dilemma,” p. 44.
11
Stecker, “The Constructivist’s Dilemma,” p. 44.
12
Joseph Margolis, “Plain Talk about Interpretation on a Relativistic Model,” The Journal of Aesthetics and Art Criticism 53 (1995): 1–7, at p. 6.
13
Stecker’s argument (“The Constructivist’s Dilemma,” p. 45) is that an interpretation which states that a work w has some representational property P, but which although reasonable and so on is not true, cannot cause w to acquire P because in order for w to be P, it must be true that w is P.
14
See, for example, H. P. Grice, “Meaning,” Philosophical Review 66 (1957): 377–388; John R. Searle, Speech Acts: An Essay in the Philosophy of Language (Cambridge University Press, 1969).
15
Philip Percival makes a similar point, although he seems to ultimately draw a different lesson from it (“Stecker’s Dilemma: A Constructive Response,” The Journal of Aesthetics and Art Criticism 58 [Note 2000] : 51–60, at p. 52).
16
Michael Krausz, “Rightness and Reasons: A Reply to Stecker,” The Journal of Aesthetics and Art Criticism 55 (1997): 415–418; Michael Krausz, “Interpretation and Its ‘Metaphysical’ Entanglements,” Metaphilosophy 31 (2000): 125–147; Percival, “Stecker’s Dilemma: A Constructive Response”; Philip Percival, “Is Constructivism Floored? Reply to Stecker,” The Journal of Aesthetics and Art Criticism 60 (2002): 82–86.
17
The reasoning underlying Percival’s strategy is a little convoluted. He argues that the very same dilemma can be raised against Stecker’s moderate intentionalist account of the meanings of artworks (Percival, “Stecker’s Dilemma: A Constructive Response,” pp. 52–53) and contends that the only way to “salvage” Stecker’s dilemma without abandoning moderate intentionalism is to find a relevant difference between interpretations and intentions (p. 53). And Percival claims the only way to do so is to take artists’ intentions regarding the meanings of their works to be truth valueless and critics’ interpretations of these artworks to have truth values.
18
Percival, “Stecker’s Dilemma: A Constructive Response,” pp. 53, 54, and 56, respectively. It is reasonably clear that, in addition to having truth values, Percival’s view is that interpretations make statements as well.
19
Percival, “Stecker’s Dilemma: A Constructive Response,” p. 60.
20
Percival, “Stecker’s Dilemma: A Constructive Response,” p. 68.
21
Percival, “Is Constructivism Floored?” p. 83.
22
Stecker, “The Constructivist’s Dilemma,” pp. 46–47.
23
Krausz characterizes realism and constructivism as exclusive and exhaustive alternatives (“Interpretation and Its ‘Metaphysical’ Entanglements,” p. 126).
24
We may have a case of apples and oranges here. Percival seems to think of interpretation types as the causes of meaning change, whereas my focus is on interpretation tokens.
25
Percival, “Stecker’s Dilemma: A Constructive Response,” p. 54.
26
For present purposes, I will ignore the question of whether the secondary encoder replaces the meaning of the work with a new meaning, supplements it with an additional meaning, or produces an entirely new work which shares a text with the original.
27
One might, of course, deny the antecedent of the conditional and claim that neither sort of meaning intention has any impact on the meaning of a work or text. This is a thorny issue, but given that the concern of this article is with giving an intentionalist defense of constructivism, rather than a more general defense of the intentionalism, it need not be addressed here.
28
Stecker, “The Constructivist’s Dilemma,” p. 44.
29
John R. Searle, “A Taxonomy of Illocutionary Acts,” in Language, Mind, and Knowledge, ed. Keith Gunderson (University of Minnesota Press, 1975), pp. 344– 369, at pp. 354–359.
30
Searle, “A Taxonomy of Illocutionary Acts,” pp. 354–355.
31
Searle, “A Taxonomy of Illocutionary Acts,” pp. 355–356.
32
Searle, “A Taxonomy of Illocutionary Acts,” p. 356.
33
Searle, “A Taxonomy of Illocutionary Acts,” pp. 356–358.
34
Of course, one’s speech act might subsequently cause a change in an artwork if, for example, one’s audience carries out a request to alter it in some way, or one subsequently fulfills a commitment to alter it oneself. The central concern here, however, is not with whether or not some kind of speech act might cause a change in an artwork, but rather whether or not the performance of a certain kind of speech act could count as an act of changing the representational properties of an artwork in its own right.
35
Searle, “A Taxonomy of Illocutionary Acts,” p. 358.
36
More recently, Stecker addresses this possibility, although not in these terms (Interpretation and Construction: Art, Speech, and the Law [Oxford: Blackwell, 2003], pp. 106– 110). Stecker’s view seems to be that insofar as a critic is making meaning declarations rather than statements, he is not engaged in interpretation. Although the constitutive conditions of interpretive discourse are beyond the scope of this article, my inclination is to suppose that meaning declarations can, in principle, in part constitute an interpretation as long as it includes statements of the requisite kinds as well.
37
Although the game might not be an authorized one in Kendall Walton’s sense. Kendall L. Walton, Mimesis as Make-Believe: On the Foundations of the Representational Arts (Harvard University Press, 1990).
38
For present purposes, I remain neutral on the question of whether in such cases the speaker performs a declaration without achieving her intended goal or fails to perform a declaration at all.
39
Jonathan Swift, Travels into Several Remote Nations of the World. In Four Parts. By Lemuel Gulliver, First a Surgeon, and Then a Captain of Several Ships (London: Benjamin Motte, 1726).
40
This is addressed in more detail below.
41
Moreover, this can occur even if, contra Margolis (“Plain Talk about Interpretation,” p. 2), there cannot be conjointly confirmed but incompatible interpretive claims if all such claims are statements and not meaning declarations.
42
Margolis, “Plain Talk about Interpretation.”
43
Stecker, “The Constructivist’s Dilemma,” p. 52, n. 12.
44
See Krausz, “Rightness and Reasons,” and Paul Thom, “Review of Rightness and Reasons, by Michael Krausz, Interpretation Radical but Not Unruly, by Joseph Margolis, and ‘The Constructivist’s Dilemma, ’ by Robert Stecker,” Literature and Aesthetics: The Journal of the Sydney Society of Literature and Aesthetics 7 (1997): 181–185, among others. Although Krausz considers imputation in some detail, he denies that he endorses it (“Rightness and Reasons,” p. 415). Nevertheless, he does seem quite sympathetic to the view.
45
Krausz, “Rightness and Reasons,” p. 417.
46
Stecker, “The Constructivist’s Dilemma,” p. 49. As I understand it, the imputation of salience to features of an object is the process by which an interpreter generates a text out of it.
47
Krausz, “Rightness and Reasons,” pp. 416–417.
48
Stecker, “The Constructivist’s Dilemma,” p. 50.
49
Krausz, “Rightness and Reasons,” pp. 417–418.
50
Thom, “Review.” It is worth noting that Thom denies that objects of interpretation generated by distinct interpretations – even those derived from a single configuration – stand in relations of numerical identity to one another. Nevertheless, distinct interpretations can be of the same thing – the further object – which they just happen to represent differently.
51
Krausz, “Interpretation and Its ‘Metaphysical’ Entanglements,” p. 141.
52
The question of whether or not texts are interpretation-dependent entities is a delicate one. Given that some of the elements that in part constitute the products of artistic activity may be conventionally or symbolically meaningful, there may exist a pattern of meaningful elements prior to and independent of anyone’s meaning intentions. On the other hand, an artist or a constructivist interpreter may mean something by additional elements that are not conventionally or symbolically meaningful or refrain from meaning anything by all of the conventionally or symbolically meaningful elements at issue. Consider, for example, a secondary encoder who means something by only every third word in a conventionally meaningful literary text. In such cases, the interpretation-generated text is distinct from what might be called the “conventional text.”
53
Now an imputationalist might well balk at my talk of interpretation-independent meanings of the elements of a work or text, insisting instead that any meanings textual elements might have are the product of one’s interpretation of the work of the whole. As Krausz puts it, “[the] interpretation prompts one to impute salience to certain features of the presented configuration, which in turn confirms the propriety of interpreting the configuration as it is” (“Interpretation and Its ‘Metaphysical’ Entanglements,” p. 138). But what we have here is not so much an argument for interpretation-dependent objects of interpretation but rather a top-down account of the meanings of artworks, that is, one according to which the meanings of textual elements are determined by the meaning of the whole work. Although I concede that there are logical relations that ideally hold between the meaning of an encoded object and the meanings of its constituent elements – which vary with the kind of representational artwork at issue – I remain neutral on the question of whether meaning is top down or bottom up.
54
One might worry that since the products of artistic activity are strictly meaningless, they could not serve as the objects of any kind of decoding interpretations. Moreover, insofar as works are text–meaning pairs, the various kind of decoding interpretations are directed toward different works. Nevertheless, if anything unifies the various projects characterized as interpretations-of-x, it is the product of artistic activity from which the various texts that in part constitute these text–meaning pairs are generated.
55
I have been assuming that (authors and) critics are sensitive to the conventional or symbolic meanings of elements of the object toward which their activities are directed, but they may, of course, disregard these as well.
Found a mistake? Contact corrections/at/cepa.infoDownloaded from http://cepa.info/3849 on 2016-11-22 · Publication curated by Alexander Riegler