Abstract
Whether “information” exists in biology, and in what sense, has been a topic of much recent discussion. I explore Shannon, Dretskean, and teleosemantic theories, and analyze whether or not they are able to give a successful naturalistic account of information—specifically accounts of meaning and error—in biological systems. I argue that the Shannon and Dretskean theories are unable to account for either, but that the teleosemantic theory is able to account for meaning. However, I argue that it is unable to account for error. Thus I conclude that while talk of informational meaning is justifiable within a naturalistic framework, talk of informational error is not, and must be used in a metaphorical sense only.
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
Notes
Clearly there are times when the translation machinery makes mistakes, and for example, the codon “CAT” which codes for “histidine” is inadvertently read as “glycine.” When this happens, it does not mean that the natural meaning of CAT is “glycine”; it just highlights the fact that the background conditions are different. It is crucial to remember that signals only carry causal information about their sources given sufficiently similar initial conditions.
Godfrey Smith (1999, p. 23) proposes that the informational content of genes does not extend past the primary structure of the proteins they code for. A gene may have many downstream effects, but that does not mean that all these effects were coded for. This can be illustrated by considering the following scenario. If one orders a pizza, this will not only cause a pizza to be made; it will also result in an oven being switched on and a cardboard pizza box being folded. Neither of these downstream effects implies that the original message contained information on box folding or oven temperatures, though. The same scenario holds for genes: they may have a number of downstream effects that contribute to the phenotype, but these effects are not coded for in the DNA; such informational content is dependent on our inferences as observers. With regards to proteins, however, that DNA contains meaningful information about their primary structure is a fact that holds independently of us.
The puzzle in the philosophy of mind is how words acquire their meaning. Philosophers of language typically say words acquire their meaning by how the speaker and/or listeners interpret them. This then passes the problem on to the philosophy of mind, for if the brain interprets words, how does it do so? There might well be brain states that mean “dog,” but in virtue of what do these neurons mean “dog”? One cannot say that certain brain states are interpreted in terms of other brain states, for therein lies a regress (Rosenberg and McShea 2008, p. 177). The teleosemantic program aims to give a fully naturalistic account of meaning, and at the same time avoid this regress (L'Hôte 2009).
This article is published in French so all page numbers correspond to the English version available online.
In terms of the philosophy of mind, the teleosemantic approach to content certainly avoids a regress. If particular neurons receive a representation, then the content of that representation is simply whatever the consumers of it have been selected to do with it. However, it is not clear that the teleosemantic approach is really explanatory in the philosophy of mind. As Devitt and Sterelny (1999, pp. 160–161) point out, most human thoughts probably have no function. And even if they did, how would such functions explain their meaning (1999, pp. 160–161)? While the situation seems much more simple in the case of DNA, it is far from clear that the functional interpretation of thoughts actually explains the contents of those thoughts. Thus the teleosemantic approach may avoid the neurological regress, but it seems to do so at the expense of being explanatory.
Maynard Smith (2000, p. 183) has suggested that DNA can be singled out as a unique informational molecule on the grounds that it represents codons symbolically, rather than indexically or iconically (in the Peircean sense). By symbolically, Maynard Smith means that there are no necessary connections between codons and their amino acid assignments. There is no chemical reason why “CAT” should code for “histidine” rather than “glycine”—the relationship is arbitrary. Whether or not the genetic code is a “frozen accident” is not the issue here; the point is that the chemical affinities between codons and amino acids play no causal role in present day protein-synthesis (Godfrey-Smith 2000, p. 204). Thus while many animal signals (e.g., pheromones and hormones) are evolutionarily contingent and can be considered arbitrary in one sense, their form is still related to their function. With DNA, however, this does not seem to be the case; it seems arbitrary in a stronger sense. Maynard Smith (2000, p. 183) notes that the way codons represent amino acids is analogous to the way that human words are related to their meanings. In linguistics, for example, the word “dog” does not naturally mean “a particular member of the genus Canis.” There are no written or phonetic properties of the word that relate it to its meaning. Instead, the content of “dog” is established by convention. This argument is not yet well worked out, but it might suggest that DNA has more semantic properties than other forms of information. If this were the case, though, it would not necessarily be philosophically problematic: representations in Millikan's triadic account can be indexical, iconical, or symbolic.
“Error” and “malfunction” will be used synonymously.
The consequences of this conclusion are actually more significant than this. Appeals to the normativity of selected functions are used in the philosophy of mind to argue for a wide variety of theses, including the norms of rationality (Nozick 1993) and the possibility of mental misrepresentation (Millikan 1984). If the etiological theory of functions is not valid, neither are any of these theories.
Davies uses the term “norm” in a normative sense here; he is not referring to the statistical norm.
Davies (2000) has also argued that even if normative functions were possessed by the type, this would still not legitimate the assignment of malfunctions. This is because functional types are defined by their selective success. In other words, functional categories are defined by those traits whose functions are undamaged and non-diseased. As soon as a trait begins to malfunction, it is disqualified from the functional category of the type. And if a trait is not part of any functional category, if it has no function, then it cannot malfunction. As Sterelny et al. (1996, p. 388) say, albeit in a slightly different context, “only things with functions can malfunction.”
References
Atkin A (2009) Peirce's theory of signs. In: Zalta EN (ed) The stanford encyclopedia of philosophy, Spring 2009 edn. http://plato.stanford.edu/archives/spr2009/entries/peirce-semiotics/. Accessed May 2010
Boorse C (2002) A rebuttal on functions. In: Ariew A, Cummins R, Perlman M (eds) Functions: new essays in the philosophy of biology and psychology. Oxford University Press, Oxford, pp 63–112
Buller D (1998) Etiological theories of function: a geographical survey. Biol Philos 13:505–527
Canguilhem G (1989) The normal and the pathological. MIT Press, Cambridge, MA
Davies PS (2000) Malfunctions. Biol Philos 15:19–38
Davies PS (2001) Norms of nature: naturalism and the nature of functions. MIT Press, Cambridge, MA
Davies PS (2009) Subjects of the world: Darwin’s rhetoric and the study of agency in nature. University of Chicago Press, Chicago
Devitt M, Sterelny K (1999) Language and reality: an introduction to the philosophy of language, 2nd edn. Blackwell, Oxford
Dretske FI (1981) Knowledge and the flow of information. MIT Press, Cambridge, MA
Dretske FI (2008) Epistemology and information. In: Adriaans P, van Benthem J (eds) Philosophy of information. Elsevier, Amsterdam, pp 29–48
Garson J (2008) Functions and teleology. In: Sarkar S, Plutynski A (eds) A companion to the philosophy of biology. Blackwell, Malden, pp 525–549
Godfrey-Smith P (1999) Genes, codes and lessons from the philosophy of mind? In: Hardcastle VG (ed) Where biology meets psychology. MIT Press, Cambridge, MA, pp 305–332
Godfrey-Smith P (2000) Information, arbitrariness, and selection: comments on Maynard Smith. Philos Sci 67:202–207
Godfrey-Smith P (2007) Information in biology. In: Hull DL, Ruse M (eds) The Cambridge companion to the philosophy of biology. Cambridge University Press, Cambridge, pp 103–119
Grice P (1957) Meaning. Philos Rev 66:377–388
Griffiths PE (1993) Functional analysis and proper functions. Brit J Philos Sci 44:409–422
Griffiths PE (2001) Genetic information: a metaphor in search of a theory. Philos Sci 68:394–403
Griffiths PE (2006) Functions, homology, and character individuation. Philos Sci 73:1–25
Kitcher P (2001) Battling the undead: how (and how not) to resist genetic determinism. In: Singh R, Beatty J, Krimbas C (eds) Thinking about evolution: historical, philosophical, and political perspectives. Cambridge University Press, Cambridge, pp 396–414
L’Hôte C (2009) Biosemantics: an evolutionary theory of thought. Evol: Education Outreach 3:265–274
Linsky B, Zalta E (1995) Naturalized Platonism vs. Platonized naturalism. J Phil 92:525–555
Maynard Smith J (2000) The concept of information in biology. Philos Sci 67:177–194
Millikan RG (1984) Language, thought, and other biological categories. MIT Press, Cambridge, MA
Millikan RG (1989) In defence of proper functions. Philos Sci 56:288–302
Millikan RG (2009) Biosemantics. In: McLaughlin B, Beckermann A, Walter S (eds) The Oxford handbook of philosophy of mind. Oxford University Press, Oxford, pp 394–406
Neander K (1991) Functions as selected effects: the conceptual analysts defense. Philos Sci 58:168–184
Neander K (2002) Types of traits: function, structure and homology in the classification of traits. In: Ariew A, Cummins R, Perlman M (eds) Functions: new essays in the philosophy of biology and psychology. Oxford University Press, Oxford, pp 390–415
Nozick R (1993) The nature of rationality. Princeton University Press, Princeton
Queiroz J, Emmeche C, El-Hani C (2005) Information and semiosis in living systems: a semiotic approach. Semiot Evol Energy Dev 1:60–90
Rea M (2002) World without design: the ontological consequences of naturalism. Oxford University Press, Oxford
Rosenberg A, McShea D (2008) Philosophy of biology: a contemporary introduction. Routledge, New York
Savitt S (1987) Absolute informational content. Synthese 70:185–190
Shannon C (1948) A mathematical theory of communication. Bell Syst Tech J 27(379–423):623–656
Shapiro S (2000) Thinking about mathematics: the philosophy of mathematics. Oxford University Press, Oxford
Shea N (2006) Millikan’s contribution to the materialist philosophy of mind. Matière Premiere 1:127–156
Shea N (2007) Representation in the genome and in other inheritance systems. Biol Philos 22:313–331
Sterelny K, Griffiths PE (1999) Sex and death: an introduction to the philosophy of biology. Chicago University Press, Chicago
Sterelny K, Smith KC, Dickison M (1996) The extended replicator. Biol Philos 11:377–403
Weaver W (1949) Recent contributions to the mathematical theory of communication. In: Shannon C, Weaver W (eds) The mathematical theory of communication. University of Illinois Press, Urbana, p 99
Wouters A (1999) Explanation without a cause. Chapter 7. Ph.D thesis, Utrecht University, p 217
Wouters A (2005) The function debate in philosophy. Acta Biotheor 53:123–151
Author information
Authors and Affiliations
Corresponding author
Rights and permissions
About this article
Cite this article
Kumar, L.A.K. Information, Meaning, and Error in Biology. Biol Theory 9, 89–99 (2014). https://doi.org/10.1007/s13752-013-0135-x
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s13752-013-0135-x