Representations as Tools
Regarding tools, it is impossible to account for their form without mention of function. Consider a simple tool like a hammer. In order to explain why hammers are the way they are, i It would be impossible to explain why a hammer is shaped as it is, or why it is not fluffier and made of cardboard without mentioning what it is for. Furthermore, it is also impossible to explain why our tools are the way they are independently of how we are ourselves. It makes no sense trying to explain why hammers have the dimensions they have without referring to the size of our arms and hands. Even if aliens developed hammers, it is very unlikely that they would be similar in shape to ours if aliens were not also similar in shape and physiology to us. In general, technology is strongly shaped, among other factors, by constraints on how users are and what they can and cannot do as much as what goals such users wish to achieve through their use of such tools. So far, this must not be controversial.
Representations are as much tools as hammers. They are also devices used to assist us in performing certain tasks. Representations help us communicate, but also help us understand and navigate the world, make art and wage war, work and play, etc. Almost every human activity exploits some form of representation to make our tasks easier, simpler ... or more fun, more beautiful, etc. As such, out theories of representation must not be much different from our theories of other human tools. In particular, in explaining why our representations are the way they are, we must take in account what we develop them for, as well as how they accommodate our human condition, our strengths and weaknesses. Just as it is a truism that a hammer for aliens might not be similar to a human hammer, so alien representations might not be similar to human representations.
Some people might object to my talk of the many uses of representation in plural. After all, it might seem that for something to be a representation, it must serve one fundamental function, i.e., to represent something. All representation are used to represent, and anything that is used with a different purpose is thus no longer a representation (This seems to be the position of David Sherry 2009, Valeria Giardino 2012 and Tarja Knuuttila 2011). The point is well taken, and so when talking about the uses of representations one must consider both their narrow function, namely to represent, and also the multiple wider purposes they serve in our lives. For example, it is true that the signs outside sex segregated public restrooms have the function of representing a man and a woman; but it must also be obvious that they are there with the purpose of helping users identify which restroom is assigned to which sex. The first is their narrow representational function, while the second is their wider purpose. Most if not all representations have a wider purpose besides their narrow representational function. Riffing on Austin’s dictum, there are many things we do with representations.
The importance of considering the wider purpose of representations, instead of focusing only on their narrow representational function must be obvious once we think of some examples. For example, it must be fairly uncontroversial to claim that any account of representational artworks that deals only with their representational function while ignoring their artistic purpose would be severely shortsighted. Something similar can be said of the images of political propaganda: they are shaped not only by their narrow representational function, but also by their wider political purposes. Failing to take both into account would leave us with a very limited picture of their nature.
Summarizing, to understand why human representations are the way they are, we must understand not only what they aim to represent, but also why we might want them to represent that and in what circumstances. Furthermore, we must remember that in order to fulfill their narrow representational function, representations must play a communicative role that essentially assigns at least two roles to their users: transmitter and receiver. In a sense, both are users of the representation as tool, and successful representations must meet the needs of both in an efficient way, that is, without spending too much of their resources. So, continuing with our previous example, in order to explain why the aforementioned restrooms signs are the way they are, we might sometimes need to appeal to their narrow representational function of representing a man and a woman, and other times to their wider purpose of helping identify restrooms for men and restrooms for women; we might sometimes need to appeal to how they fit the needs, limitations, etc. of potential restrooms users who can benefit from the information contained in the signs, and other times to those of whoever wants to communicate that information to them. For example, we might explain why the figures in the women’s restrooms signs are wearing a skirt by appealing to how, even though neither all women wear skirts, nor everyone who wears a skirt is a woman, we (possible restroom users) expect skirt wearers to be women, and thus we (who want to help these potential users identify the sex corresponding to each restroom) can exploit this expectation to signal a sexual difference between these figures and the figures in the other restroom.
A lot of philosophy, specially in the philosophy of science has focused on a particular subset of representations: those that are used as inferential tools, that is tools hat aid us in making inferences. When dealing with inferential representations, therefore, we must distinguish between the question of how our inferential goals shape the representations we use to assist us in such inferences, and the question of how such representations complement, exploit, and extend our cognitive abilities. Consequently, when dealing with successful inferential representations, the questions of why we use them entails two different but closely related questions: first, why it is advantageous for us to use them, and second, why are we justified in using them. In other words, we must ask both how well the representation fits the task, and how well it fits us; we must ask both how efficient it is to use the representation, and how efficacious it is. Call questions about how our constitution as users shape our inferential representations “ergonomic” to distinguish them from more “logical” or “epistemological” questions about their inferential efficacy. Needless to say, philosophers have focused almost exclusively on these later questions and have thus ignored an important dimension of our inferential representations.