The position that one constructs a literary work when one reads it is commonly called constructivism. Robert Stecker has defined constructivism in this way: "the idea that interpretations are not simply instruments for discovering properties already to be found in works, but contribute to the creation of such properties." (96) (Interpretation and Construction: Art, Speech, and the Law, Blackwell, 2003). So the question is whether the properties are already there or are interpretations such that they also contribute to the creation of properties (i.e. is the creative process of a work of art such that it continues beyond the completion of the work by the artist...continues, that is, in the process of understanding and interpretation of the work as well as in various responses to the work in various media as for example in different productions of a play). I have argued elsewhere (in a forthcoming article on literary interpretation) that interpretations actualize properties that exist only as potential and which may be actualized in many ways. So I suppose that makes me a constructivist.
Theodore Gracyk in The Philosophy of Art (Polity, 2012) argues against constructivism. Interestingly, he situates this debate in the context of a discussion of authenticity and cultural origins (Chapter 5 of his book.) As a result, he sees constructivism as problematic because it leads to culturally inauthentic use of artworks. Gracyk also bases his attack on constructivism on a distinction between properties of an artwork that are essential and ones that are contingent: on this view, constructivists take contingent properties and treat them as though they were essential. The issue, of course, is whether the meaning of a literary or other work of art is frozen in time at the moment of its completion (as ontological contextualists like Gracyk would argue) or whether they change over time, taking on new possibilities and hence new, although often very different, interpretations (all of which have their own value.) An example of a contingent property would be the name of Velazquez' painting Rokeby Venus: it was named this way because it was associated with the Rokeby Hall, but could have been named something else if it was associated with another location. That it is called a Venus however would be essential, for Velazquez surely intended it to represent the goddess Venus. It is notoriously difficult to determine which properties are essential, for example is it essential to this painting that it was painted by Velazquez? On the other hand some properties are obviously not essential, for example that Gracyk has seen the painting.
The question that determines whether a property is essential is traditionally "could one imagine the same object without the property in question?" This is a peculiar question and it is hard to know how to answer it in many cases since it depends on what one's criteria of sameness is as well as how good one is at imagining. Although philosophers tend to ask this question as though the answer must be the same for everyone, it never is. So one suspects that it only tells us what is essential for this person or for this group, rather than what is essential as such.
Gracyk favors a view called "ontological contextualism." This view (as initially described) is that "some aspects of an artworks identity depend on the art-historical context of its time of creation." It seems hard to argue against this. However, sometimes ontological contextualism really amounts to the view that all aspects of the artworks identity depend on art-historical context of the time of creation. This is the view that constructivists deny. It is also the view which Gracyk defends, for he also says that "Ontological contextualists owe constructivists a clear account of why earlier art-historical contingencies are relevant to artwork identity, yet later ones are always irrelevant." (my italics, 86)
Gracyk also says that "according to constructivism, there are no good reasons to 'freeze' identity at the time of creation...Artworks gain and lose properties after their creation, reflecting the art-historical context of audience appreciation and interpretation." (85) This seems basically right to me, but then I am a constructivist. Gracyk mentions two arguments offered in favor of constructivism: that later artworks can reorder the importance of properties of earlier artworks, and that meanings of words can change. The second view need not detain us since it doesn't make sense to interpret a poem that uses the term "gay" in the sense of "colorful" or "happy" to mean "gay" in the sense of "homosexual" when that word did not have that meaning at the time of the creation of the work. The first point however is telling in support of constructivism. The claim against this by the ontological constructivist then is based on this sort of assertion: "Cezanne's painting were always proto-Cubist. It simply took the emergence of full-blown Cubism to reveal this aspect of Cezanne's style." On this view, the future is always already in the past. One wonders, however, how a claim like this could ever be supported. It must rather be an item of faith! We see something differently because of new developments and it can inevitably be said that this aspect was always already there every bit as much as Jupiter's moons were always there before the invention of the telescope. The analogy just doesn't seem plausible. We know that Jupiter's moons were there before the telescope was invented because of other correlating evidence: the same cannot be said for the Cubist aspect of Cezanne's paintings!
Another argument against constructivism offered by Gracyk is that it "is only plausible so long as we regard every artwork as an abstract structure that lacks determinate meaning." On this view, constructivists believe that the essential properties of a work can change over time as long as one changes the symbol system by which one interprets the work. But, Gracyk argues, just because we have two texts that are identical we do not necessarily have the same work: two identical texts can come out of different times and have different meanings. I am not convinced that a constructivist must view artworks as abstract structures. Artworks have meanings and interpretation of them is conditioned in various ways: that they were produced at particular times and by particular persons is important, but this does not mean that they are immune to change: as Heraclitus taught us, nothing is immune to change. Artworks are not abstract structures but rather are nests of potentiality which may be actualized into determinate meaning at different times in different ways by different interpreters, and in an historically evolving sequence. The identical text scenario, then, does not apply to the issue at hand.
Gracyk's most interesting argument for ontological contextualism is in terms of cultural appropriation. Ontological contextualism holds that "in object appropriation [where a physical object is removed from its cultural provenance, for example Picasso's Guernica to New York MOMA] the essential properties of an object are exactly the same before and after its appropriation." On the other hand, in the case of design appropriation [when a design originating in one culture is used by another culture] "the appropriation alters the meaning of what is transferred." (89) So, on this view, contextualism works as long as one is talking about copies and not originals: "the lesson of ontological contextualism is that the copy has essential, non-manifest properties that differ from those of the source object or design." This leads Gracyk to argue that design appropriation is not theft because it does involves alteration of identity. Does this pose a serious problem for constructionism? Gracyk agrees that the copy and its use can be understood in terms of constructionism, but not the original. Is the Picasso the same in New York as it was in Spain or France? Is it just contingent that it is in New York? Is its New York presence irrelevant to its interpretation? Do I actualize it differently when I see it at the MOMA? Do we have a clear idea of what is contingent here and what is essential?
Gracyk's most powerful argument against constructivism is moral: constructivism blurs the distinction between stable, essential properties that depend on provenance and properties that vary with audience perspective" for "it implies that 'outsider' responses to the artifacts and artworks of unfamiliar cultures cannot be dismissed as misinterpretations." He then insists that the constructivist believes that "properties are imposed rather than discovered during the process of interpretation." (91) But there is nothing about the definition of constructivism that insists on this point or on the imposed/discovered dichotomy. The notion that properties are imposed by interpretation implies a kind of cultural dictatorship...it is very negative sounding. One could better say that the constructivist believes that properties are actualized in different ways rather than simply discovered in the way the moons of Jupiter were discovered. Note also, that Gracyk had already considered that in the case of copying designs, the outsider response cannot, in this case, be dismissed as a misinterpretation. Let's just say that outsider responses (and even insider responses) can sometimes be seen as misinterpretations, and even dismissed as really bad misinterpretations, but that this is not always required or recommended.
Gracyk thinks that when a non-Navajo imitates a Navajo sand painting and thus strips away the ritual significance of the work then the identity is altered. But the identity of the original sand paintings is not altered, so what sense can be made of the claim? One can say that there are ethical problems with such imitations, but it is not clear that constructionism authorizes this activity in a way the ontological contextualism does not. I agree that many cultural appropriations are offensive, and that we have moral obligations in this regard. Moreover, Gracyk is certainly right that "the non-Navajo who fantasizes that appropriation of a Navajo design preserves and respects Navajo spirituality is engaged in wishful thinking." (92) How is this a problem for constructivism?
Gracyk's main point is that "ontological contextualism recognizes that artwork identity is a matter of cultural provenance" and constructivism does not, largely because it confuses issues of identity by dissolving or at least messing with the distinction between essential and accidental properties. My response is that constructivism need not deny that identity is a matter of cultural provenance, that the very notion of "essential properties" is vague and confusing, and that important properties of works evolve and shift over time often following a historical or even evolutionary path where different valid interpretations are allowable as different contexts emerge, context not limited to a moment frozen in time.