The received view holds that there is a significant divide between full-blown representational states and so called ‘detectors’, which are mechanisms set off by specific stimuli that trigger a particular effect. The main goal of this paper is to defend the idea that many detectors are genuine representations, a view that I call ‘Strong Liberal Representationalism’. More precisely, I argue that ascribing semantic properties to them contributes to an explanation of behavior, guides research in useful ways and can accommodate misrepresentation.

Representations may have descriptive content, directive content, or both, but little explicit attention has been given to the problem of distinguishing representations of these three kinds. We do not know, for instance, what determines whether a given representation is a directive instructing its consumer to perform some action or has descriptive content to the effect that the action in question has a certain value. This paper considers what it takes for a representation to have directive content. The first part of (...) the paper presents the Liberal View, which might be taken to be the default position on this issue. The Liberal View has some attractions, but as the second part shows, these are less conclusive than they might at first appear, and there is much to be said for an alternative, the Strict View. (shrink)

How can we think about things in the outside world? There is still no widely accepted theory of how mental representations get their meaning. In light of pioneering research, Nicholas Shea develops a naturalistic account of the nature of mental representation with a firm focus on the subpersonal representations that pervade the cognitive sciences.

One of the central tasks for naturalistic theories of representation is to say what it takes for something to be a representation, and some leading theories have been criticised for being too liberal. Prominent discussions of this problem have proposed a producer-oriented solution; it is argued that representations must be produced by systems employing perceptual constancy mechanisms. However, representations may be produced by simple transducers if they are consumed in the right way. It is characteristic of representations to be consumed (...) by systems which are capable of independent action. This paper defends this claim; discusses more precise, naturalistic formulations; and shows how it can illuminate the explanatory payoffs which science achieves by appealing to representation. (shrink)

Desires, or directive representations, are central components of human and animal minds. Nevertheless, desires are largely neglected in current debates about the naturalization of representational content. Most naturalists seem to assume that some version of the standard teleological approach, which identifies the content of a desire with a specific kind of effect that the desire has the function of producing, will turn out to be correct. In this paper I argue, first, that this common assumption is unjustified, since the standard (...) approach is in fact deeply problematic. Secondly, I propose an alternative account of the content of desires which, while generating plausible and determinate content ascriptions, avoids the main problem that plagues the standard approach, and is also preferable on other grounds. (shrink)

In 1965, Konrad Lorenz grounded the innate–acquired distinction in what he believed were the only two possible sources of information that can underlie adaptedness: phylogenetic and individual experience. Phylogenetic experience accumulates in the genome by the process of natural selection. Individual experience is acquired ontogenetically through interacting with the environment during the organism’s lifetime. According to Lorenz, the adaptive information underlying innate traits is stored in the genome. Lorenz erred in arguing that genetic adaptation is the only means of accumulating (...) information in phylogenetic experience. Cultural adaptation also occurs over a phylogenetic time scale, and cultural tradition is a third source from which adaptive information can be extracted. This paper argues that genetic adaptation can be distinguished from individual and cultural adaptation in a species like Homo sapiens, in which even adaptations with a genetic component require cultural inputs and scaffolding to develop and be expressed. Examination of the way in which innateness is used in science suggests that scientists use the term, as Lorenz suggested, to designate genetic adaptations. The search for innate traits plays an essential role in generating hypotheses in ethology and psychology. In addition, designating a trait as innate establishes important facts that apply at the information-processing level of description. (shrink)

Ethological theories usually attribute semantic content to animal signals. To account for this fact, many biologists and philosophers appeal to some version of teleosemantics. However, this picture has recently came under attack: while mainstream teleosemantics assumes that representational systems must cooperate, some biologists and philosophers argue that in certain cases signaling can evolve within systems lacking common interest. In this paper I defend the standard view from this objection.

One of the main tenets of current teleosemantic theories is that simple representations are Pushmi-Pullyu states, i.e. they carry descriptive and imperative content at the same time. In the paper I present an argument that shows that if we add this claim to the core tenets of teleosemantics, then (1) it entails that, necessarily, all representations are Pushmi-Pullyu states and (2) it undermines one of the main motivations for the Pushmi-Pullyu account.

Ethological theory standardly attributes representational content to animal signals. In this article I first assess whether Ruth Millikan’s teleosemantic theory accounts for the content of animal signals. I conclude that it does not, because many signals do not exhibit the required sort of cooperation between signal‐producing and signal‐consuming devices. It is then argued that Kim Sterelny’s proposal, while not requiring cooperation, sometimes yields the wrong content. Finally, I outline an alternative view, according to which consumers alone are responsible for conferring (...) representational status and determining content. I suggest that consumer‐based teleosemantics reconstruct the content of both cooperative and noncooperative signals and explain how a given trait can mean different things to different consumers. †To contact the author, please write to: Department of Philosophy, King’s College London, Strand, London WC2R 2LS, U.K.; e‐mail: [email protected]. (shrink)

There is ongoing controversy as to whether the genome is a representing system. Although it is widely recognised that DNA carries information, both correlating with and coding for various outcomes, neither of these implies that the genome has semantic properties like correctness or satisfaction conditions, In the Scope of Logic, Methodology, and the Philosophy of Sciences, Vol. II. Kluwer, Dordrecht, pp. 387–400). Here a modified version of teleosemantics is applied to the genome to show that it does indeed have semantic (...) properties – there is representation in the genome. The account differs in three respects from previous attempts to apply teleosemantics to genes. It emphasises the role of the consumer of representations. It rejects the standard assumption that genetic representation can be used to explain the course of an organism’s development. And it identifies the explanatory role played by representational properties of the genome. A striking consequence of this account is that other inheritance systems could also be representational. Thus, a version of the parity thesis is accepted. However, the criteria for being an inheritance system are demanding, so semantic properties are not ubiquitous. (shrink)

This paper investigates the prospects of Rodney Brooks’ proposal for AI without representation. It turns out that the supposedly characteristic features of “new AI” (embodiment, situatedness, absence of reasoning, and absence of representation) are all present in conventional systems: “New AI” is just like old AI. Brooks proposal boils down to the architectural rejection of central control in intelligent agents—Which, however, turns out to be crucial. Some of more recent cognitive science suggests that we might do well to dispose of (...) the image of intelligent agents as central representation processors. If this paradigm shift is achieved, Brooks’ proposal for cognition without representation appears promising for full-blown intelligent agents—Though not for conscious agents. (shrink)

Teleosemantic theories of representation are often criticized as being “too liberal”, i.e. as categorizing states as representations which are not representational at all. Recently, a powerful version of this objection has been put forth by Tyler Burge. Focusing on perception, Burge defends the claim that all teleosemantic theories apply too broadly, thereby missing what is distinctive about representation. Contra Burge, I will argue in this paper that there is a teleosemantic account of perceptual states that does not fall prey to (...) this problem, and that we can arrive at this account by combining some of Burge’s insights with a producer-oriented version of teleosemantics. The resulting theory turns out to be attractive and perfectly coherent. By contrast, the coherence of Burge’s own anti-teleosemantic approach becomes quite doubtful under closer examination—or so I will argue. (shrink)

The success of a piece of behaviour is often explained by its being caused by a true representation (similarly, failure falsity). In some simple organisms, success is just survival and reproduction. Scientists explain why a piece of behaviour helped the organism to survive and reproduce by adverting to the behaviour’s having been caused by a true representation. That usage should, if possible, be vindicated by an adequate naturalistic theory of content. Teleosemantics cannot do so, when it is applied to simple (...) representing systems (Godfrey-Smith 1996). Here it is argued that the teleosemantic approach to content should therefore be modified, not abandoned, at least for simple representing systems. The new ‘infotel-semantics’ adds an input condition to the output condition offered by teleosemantics, recognising that it is constitutive of content in a simple representing system that the tokening of a representation should correlate probabilistically with the obtaining of its specific evolutionary success condition. (shrink)

According to Karen Neander’s causal-informational teleosemantics, the contents of perceptual states depend on the etiological response functions of sensory-perceptual systems. In this paper, I argue that this theory is, despite its virtues, unable to explain how humans and other animals are capable of perceiving properties with which no sensory-perceptual system has ever been confronted. After rejecting Neander’s own proposal in terms of second-order similarity and a proposal inspired by Ruth Millikan in terms of simplicity, I offer a solution which equates (...) functions with manifestations of dispositions that made a difference to evolutionary success. My suggestion is able to generate determinate and plausible contents for new perceptual states while still preserving the idea that etiological functions explain evolutionary success. (shrink)

One way to defend humane animal agriculture is to insist that the deaths of animals aren’t bad for them. Christopher Belshaw has argued for this position in the most detail, maintaining that death is only bad when it frustrates categorical desires, which he thinks animals lack. We are prepared to grant his account of the badness of death, but we are skeptical of the claim that animals don’t have categorical desires. We contend that Belshaw’s argument against the badness of animal (...) death relies on overly simplistic thought experiments and isn’t sufficiently careful in how it attributes mental states to animals. We present some cases of animal behavior from recent work on animal cognition that are most plausibly understood as spurred by categorical desires and we show how an independently plausible account of mental content—Ruth Millikan’s teleosemantics—supports our attribution of categorical desires in those cases. We end by arguing that even if you’re wary of our reliance on teleosemantics, you should still accept it due to considerations about moral risk. (shrink)

This paper explores John Maynard Smith’s conceptual work on animal signals. Maynard Smith defined animal signals as traits that (1) change another organism’s behaviour while benefiting the sender, that (2) are evolved for this function, and that (3) have their effects through the evolved response of the receiver. Like many ethologists, Maynard Smith assumed that animal signals convey semantic information. Yet his definition of animal signals remains silent on the nature of semantic information and on the conditions determining its content. (...) I therefore compare three ways to specify the semantic content of animal signals. The first suggestion models semantic content on Maynard Smith’s theory of genetic information. On the second proposal, semantic content is equated with a condition identified by conventional content ascriptions. The third suggestion is to explain semantic content in terms of consumer-based teleosemantics. I show how these accounts equate semantic content with distinct kinds of conditions and how they differ with respect to the kinds of traits that qualify as carrying semantic information. (shrink)

There are many different notions of information in logic, epistemology, psychology, biology and cognitive science, which are employed differently in each discipline, often with little overlap. Since our interest here is in biological processes and organisms, we develop a taxonomy of functional information that extends the standard cue/signal distinction (in animal communication theory). Three general, main claims are advanced here. (1) This new taxonomy can be useful in describing learning and communication. (2) It avoids some problems that the natural/non-natural information (...) distinction faces. (3) Functional information is​ ​produced through exploration and stabilisation processes. (shrink)

We argue that Brette's arguments, or some variation on them, work only against the immodest codes imputed by neuroscientists to the signals they study; they do not tell against “modest” codes, which may be learned by neurons themselves. Still, caution is warranted: modest neural codes likely lead to only modest explanatory gains.