Reference, Ontology, and Architecture: Response to Rafael De Clercq

Reference, Ontology, and Architecture: Response to Rafael De Clercq Appreciative practices vary across cultures, and some of these variations are philosophically interesting. A case in point is one strand of Japanese architectural appreciation, strikingly illustrated by the Shinto shrine Ise Jingu. Japanese thinking about this shrine appears to contrast with thinking about architecture in the West. In “Shikinen Sengu and the Ontology of Architecture in Japan,” I proposed an ontological hypothesis to explain the contrast.1 Rafael De Clercq suggests an alternative, linguistic hypothesis that is both enlightening and plausible. However, a case can be made that both hypotheses are true if the linguistic hypothesis is true. To see how some Japanese thinking about architecture contrasts with thinking about architecture in the West, consider the shikinen sengu. This is the practice, dating back many centuries, of dismantling the core structures at Ise Jingu and then ritually rebuilding them from scratch every twenty years. Japanese discourse about the shrine includes the assertion of the following pair of propositions. (O) Ise Jingu is more than twenty years old. (Y) Ise Jingu is no more than twenty years old. These propositions are contradictory unless ‘Ise Jingu’ is ambiguous. Reading them univocally, Western thinking about architecture endorses (Y) and denies (O). Two questions arise off the bat. First, how might we disambiguate ‘Ise Jingu’ to make sense of the joint assertion of (O) and (Y)? Second, what lies behind the choice in the West of (Y) over (O)? According to the ontological hypothesis, ‘Ise Jingu’ in (O) and (Y) designates entities of different ontological kinds. What these different kinds are is an open question. Maybe ‘Ise Jingu’ refers to a type in (O) and to a material object that is a token of that type in (Y); or maybe it refers to an ongoing ritual event in (O) and a part of that event in (Y). The ontological hypothesis is a prompt to further research. What about the second question? Why choose (Y) over (O)? One answer is that architectural works are mostly taken in the West to be material objects, on the commonsense understanding of material objects. On this understanding, material objects have persistence conditions that imply that no material object survives the simultaneous replacement of all its parts. Call the structures that stand at Ise Jingu before and after the last sengu “A” and “B” respectively. Since all the parts of A are simultaneously replaced with parts of B, A is not B. So if Ise Jingu is a material object on the commonsense understanding of material objects, (Y) is true and (O) is false. De Clercq proposes a linguistic hypothesis to answer our two questions. The hypothesis is that in Japan ‘Ise Jingu’ is a temporally nonrigid designator like a definite description. Since it is temporally nonrigid, it may refer to different objects at different times. Thus ‘Ise Jingu’ refers to A before the last sengu and to B afterward. So (Y) is true if ‘Ise Jingu’ designates B and B is not more than twenty years old. And (O) is true if more than twenty years ago there was an item designated by ‘Ise Jingu.’ Turning to the second question, De Clercq suggests that the habit in the West is to treat names of architectural works as temporally rigid partially descriptive names—as names that refer to the same object at all times when the object exists. Treating ‘Ise Jingu’ as such a name means that ‘Ise Jingu’ is one name when it refers to A and it is another name when it refers to B, not one name that has had a referent for centuries. So (O) is false. There is some independent evidence in favor of the linguistic hypothesis. Saul Kripke famously argued that proper names are not synonymous with descriptions. Suppose that Schmidt actually proved the incompleteness theorem and Gödel got hold of and claimed credit for the work. If ‘Gödel’ means something like “the man who proved the incompleteness theorem,” then in using the name ‘Gödel’ we would refer to Schmidt. But our intuitions tell us that we would not refer to Schmidt in using the name ‘Gödel.’ So ‘Gödel’ does not mean something like “the man who proved the incompleteness theorem.”2 Kripke's argument appeals to intuitions, which presumably reflect our referential practices. However, recent empirical research suggests that not everybody shares these intuitions, and referential practices vary culturally.3 North American and East Asian subjects were presented with a version of Kripke's story about Gödel and Schmidt, and they were asked whether a person who uses the name ‘Gödel’ is talking about (a) the person who proved the incompleteness theorem or (b) the person who got hold of the work and claimed credit for it. Although North Americans tend to follow Kripke in answering (b), East Asians tend to answer (a). The authors of the study conclude that East Asians “tended to have descriptivist intuitions, while Westerners tended to have Kripkean ones.”4 Needless to say, not much can rest on a single study—the results must be duplicated and the methodology validated—but if the study is correct, then it supports the linguistic hypothesis. Is the ontological or the linguistic hypothesis correct? The correct hypothesis is the one that best explains the contrasting features of practices of architectural appreciation in Japan and the West. However, the hypotheses are not contraries, and a case can be made that both are true if the linguistic hypothesis is true. De Clercq compares ‘Ise Jingu’ with ‘the president of the United States’: both are temporally nonrigid descriptive designators. Thus (O) and (Y) parallel: (OP) the president of the United States has lived in the White House for more than a hundred years. (YP) the president of the United States has lived in the White House for less than eight years. So (YP) is true if ‘the president of the United States’ designates George W. Bush and he did not live in the White House more than eight years ago. And (OP) is true if more than a hundred years ago there lived in the White House an item designated by ‘the president of the United States.’ Now consider the following: (OR) the fastest person in the world has lived in Kenya for two hundred years. (YR) the fastest person in the world has lived in Kenya for less than fifty years. Again, if (OR) is true, then ‘the fastest person in the world’ has had a referent for centuries. Obviously, though, the (OP)/(YP) pair differs significantly from the (OR)/(YR) pair. Whereas ‘the president of the United States’ refers to an office in (OP), in (OR), ‘the fastest person in the world’ refers to nothing like an office. That is, there are facts about the subject of (OP) that are not facts about the subject of (YP).5 For example, it is true that the president of the United States defends the Constitution, but it is false that the forty‐third president of the United States defends the Constitution. The question then arises whether (O) and (Y) are better understood on the model of (OP) and (YP) or (OR) and (YR). If (O) and (Y) are more like (OR) and (YR), then De Clercq is right and the linguistic hypothesis displaces the ontological hypothesis. A closer examination of discourse in Japanese about Ise Jingu and other architecture might vindicate this scenario. But if (O) and (Y) are more like (OP) and (YP), then perhaps the linguistic hypothesis stands alongside the ontological hypothesis. According to the latter, ‘Ise Jingu’ refers in (Y) to a material object and refers in (O) to something other than a material object. In effect, the descriptive content of ‘Ise Jingu’ involves different sortal concepts in (O) and (Y).6 In (Y), the sortal concept yields the identity and persistence conditions of a material object. The sortal concept associated with ‘Ise Jingu’ in (O) yields different identity and persistence conditions. If a referent of ‘Ise Jingu’ has aesthetic properties in virtue of its having been around for centuries, then that referent is not identical to any building taken down after twenty years. For example, the referent of (O) represents a cycle of decay and rebirth—but not the referent of (Y). As the point is often put, the former is a work in its own right. If this work has different identity and persistence conditions than material objects, then the ontological hypothesis is true. As always, the place to look to settle the matter is our appreciative practices. They can tilt us toward the ontological hypothesis, the linguistic hypothesis, or a conjunction of the two. The linguistic hypothesis displaces the ontological hypothesis if (O) and (Y) are like (OR) and (YR). The hypotheses are compatible if (O) and (Y) are like (OP) and (YP) and ‘Ise Jingu’ is a descriptive designator. Finally, if ‘Ise Jingu’ is used as a Kripkean name in Japan, then the ontological hypothesis is our only option. In the end, all options challenge some widespread philosophical doctrines about the cultural invariance of ontology or reference.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.