Abstract

We represent the world with subject-predicate representation, referring to an object and attributing a property to it. But could it be otherwise? Could other, alien languages do things very differently? Could there be languages that have very different semantic categories, and with it, could the world have a very different structure than what we take it to be: a world not of objects having properties but some other, alien structure?How serious would it be if the answer were yes? Maybe not that serious. Maybe there is more than one way to describe the world, with objects having properties being one of them, but not the only one. If so, then nothing about our representations of the world is challenged; they just aren’t unique. But maybe it is very serious. Maybe these different ways are in conflict, and only one of them can be right. If so, then maybe only the alien way of representing the world gets it right and all our efforts are in vain. In that case the world might be very different than how we represent it to be, maybe even ineffable for us and beyond what we can represent, at least if we are stuck with subjects and predicates. But if we can go beyond it, we might need to find completely different languages to do better. Should we take this possibility of alien languages and alien structure seriously?In his terrific new book, Matti Eklund argues that we should. In doing so he goes over a series of topics and debates in recent metaphysics and the philosophy of language that are connected to this issue. As Eklund repeatedly states, his aims in this book are modest in that all he tries to do is to show that we should take this possibility very seriously. In particular, he wants to show that arguments against this possibility are mistaken, but also that alleged examples of alien languages are mistaken as well. The body of the book consists of a discussion of such topics and arguments, with chapters structured by groupings of them according to some general theme: ontological categories, somewhat alien languages, metaphysical theorizing without objects and properties, representing the world in ways other than language, and so on and so forth. In doing this Eklund covers many of the main recent topics and themes in the intersection of metaphysics and the philosophy of language. To name just a few, we encounter quantifier variance, neo-Fregean conceptions of objects and their relation to singular terms, metaontology, ontological nihilism, the philosophy of physics, feature-placing languages, and many more. The discussion considers questions such as: Are there human languages that are alien relative to English? Does work in the philosophy of physics support the idea that our present semantic categories are inadequate to describe physical reality? Are there coherent metaphysical systems that dispense with objects and properties? Are maps distinctly different from natural languages in how they represent the world? Do various positions in metaontology support or undermine alien structure? Can we make sense of there being something that isn’t a thing? Could there be a language without predication, and is there evidence that some actual human languages lack predication? And so on and so forth. It is a true tour de force of discussions of topics related to alien languages and structure. Eklund has himself made important contributions to many of these topics over the last few decades, and the discussion strikes me as excellent throughout. It is not only beautifully set up and persuasively argued but also an expert survey of many of the main issues in the field and their connection to the topic of alien structure. I found myself in agreement with many of the arguments concerning whether a particular position does or does not show that alien languages and structure are possible, except for when my own views were criticized. That this discrepancy is food for thought is not lost on me.As I mentioned, Eklund himself takes the aim of his book to be modest: to argue that alien structure should be seen as possible. He does not think that any of the arguments he considers properly settle this question; rather, he thinks that they overall support the view that alien structure should be seen as possible, since, as he puts it, “Why not?” (244).I don’t want to quibble with the discussion that Eklund engages in about the many arguments and topics in this area, in part because I often agree and in part because I do feel a more urgent sense of disagreement about the overall setup of the book. There are really two things that I feel some unease about, and I would like to focus the more substantial critical parts of this review on them. The first concerns the role that our own language plays in setting up this issue. The second concerns the relationship between alien structure and ineffability.When we consider alien languages and alien structure, there are really two ways of approaching this issue. The first is to step back and think of our language and familiar structure next to an alien language and alien structure and to compare them from a top-down vantage point. Here are two options, so why this one rather than that one? That is in essence Eklund’s approach. A second option is to carefully consider the language you use to state the problem and then to evaluate alternatives in light of the fact that you used a particular language to state the very problem you hope to solve. That can make a difference, or so it seems to me. To illustrate it, consider Eklund’s frequent talk of representing the world with this or that language. Might the world be better represented with an alien language that doesn’t engage in reference to objects and attributes properties? But what is ‘the world’ in this setup? One way to think about it is this: The world is just reality. Reality, in turn, famously can be thought of as either the totality of things or the totality of facts. Let’s take the former for the moment. If the world is the totality of things, then obviously the world can’t be properly represented in a language that doesn’t have the capability to refer to things or something equivalent. A truly alien language that has nothing to do with reference to things won’t do great at representing the world, understood as the totality of things. And similar remarks hold for thinking of the world as the totality of facts. Setting the issue up by asking about how we might alternatively represent the world sets a certain goalpost that is determined by the language we used in formulating the problem. And that language is our familiar language.This issue does not really play much of a role in Eklund’s discussion, although an analogous one was more prominent in his last book, Choosing Normative Concepts (2017), which concerned which normative concepts we should have. Asking this question this way, using “should” in the question, might seem to favor an answer that it should be should and not some alternative, alien normative concept should*. The same general issue can be seen as also applying to the present book on alien structure. If we wonder about the structure of the world, then maybe that favors an answer where the world has object-property structure, or the structure to be found in facts. But if we asked about something else, the structure of some alien substitute for the world, world*, then things might be different. Still, we asked about the world, and with that goalpost set, does it not favor the familiar over the alien?This might seem to push the issue back to why we should set the goalpost the way we did, asking about alternative ways to represent the world. Why not world*? But be that as it may, as the question was asked, the answer seems to favor the familiar over the alien. And asking about the world is not just some arbitrary topic; it is what we are trying to find out about in general. Maybe setting the goal this way is halfway there to reaching it, or maybe it itself needs to be reevaluated. But from where and with what language would we do this? You can’t just start from nowhere, using no language to reflect on which language should be used to ask the question we begin with. So, if our starting point is fine and the question as it was asked is acceptable, doesn’t the familiar win almost by default?Maybe not, and this brings me to my second main point about Eklund’s overall discussion. Even if we focus our question on the world, there might be parts of the world that are alien and others that are not, as Eklund discusses. If the world is understood as the totality of facts, then the issue of alien structure quickly turns into an issue of ineffability. If there are facts that require a representation with a structure completely different than the structures our present languages have, then such facts cannot be represented in our present languages and thus are ineffable at present. If this limitation is tied to a contingent feature of our languages, then maybe it can easily be overcome, adding the required resources. But if the limitation is more intimately tied to our mental architecture, then maybe it can’t be overcome. It could be that there are some facts that require a representation with a structure so alien to our minds that we just can’t represent the world in this way. Thus, structure can be a source of ineffability. We can call a fact structurally ineffable if representing it requires a representation with a structure we can’t employ. It would simply be beyond what we can represent with minds like ours. I would argue that it is this connection that makes wondering about structure especially significant. Ineffability is what ultimately matters, structure alone is secondary. If structure is divorced from ineffability, then it is still interesting, since it would highlight that the same fact can have different structures and thus can be represented with very different kinds of representations. But if alien structure is real and gives rise to ineffability, then this makes all the difference. Structural ineffability would challenge how creatures like us can carry out metaphysics at all and how we can hope to come to an understanding of reality as a whole. It would also raise the issue why our representational limitations and structural ineffability have so far not played a major role in our attempts at inquiry. If structural ineffability is real, shouldn’t it play a more prominent role? But if ineffability is ruled out, then this also would have major implications and give rise to speculation about why we just happen to have all the structure in our language that allows us to represent the structure in all the facts. Maybe reality is not independent of our conceptualization of it, maybe there is some other explanation of it, or maybe we just got lucky. In any case, I find the focus on ineffability as the ultimate topic more apt than focusing on structure.1 Eklund is less impressed with ineffability, suggesting that it likely can be overcome with reference to sentences in alien languages that express facts that we can’t directly express in our language. I have my doubts about that this resolves the issue, but either way, Eklund’s focus on structure and not ineffability has significant consequences for his overall presentation of these issues.I would like to highly recommend this excellent book to anyone interested in these foundational issues about our language and the world. It is asking an important question, it gives a great overview of related issues, and it is expertly presented and argued, but it is also modest in that it focuses on whether alien structure is a real possibility, not so much on what would follow if it were or weren’t. That must be the next question now, and I am confident that this book will spark much great work to follow on this central topic.