The Daily Ant hosts a weekly series, Philosophy Phridays, in which real philosophers share their thoughts at the intersection of ants and philosophy. This is the forty-fifth contribution in the series, submitted by Kelle Dhein.
How Ants Can Help Solve the Mystery of Intentionality
Biologists ascribe meaning to living systems all the time. Geneticists are happy to say that DNA carries information about how to build an organism. Bee researchers claim the waggle dance communicates information about the location food to nest mates. And as two Philosophy Phriday contributors have already noted (Lorraine Keller and Kevin Lande), the impressive navigational abilities of the desert ant Cataglyphis have caused researchers to claim that the ant must utilize cognitive representations in some way.
What all those claims have in common is that they posit an intentional relationship between a living system and some other thing. “Intentionality” is a philosopher’s term for meaning, reference, representation, and general about-ness. Although the English word “intention” seems to form the root of “intentionality”, philosophers use intentionality as a technical term with no necessary connection to willful intentions. For example, a map of the Grand Canyon seems to exhibit intentionality in relation to the actual Grand Canyon in Arizona, not because the map somehow intends to represent the Grand Canyon but because the lines on the map are structured in such a way that they correspond to real features of the Grand Canyon.
Philosophers have yet to produce a consensus account of how intentionality works, so when philosophers see biologists positing intentional relationships in the living world, they generally ask two questions: (1) How are biologists using intentional concepts? (2) How does that usage square with existing philosophical work? In the case of desert ants integrating information about distance traveled, the answers are far from clear.
One possible answer comes from the work of philosopher Ruth Millikan, whose teleosemantic theory of intentionality holds that intentional relationships supervene on evolutionary history.1 The idea is that things in the biological world enter into intentional relationships with each other by virtue of the biological functions they have mediated in the past. For example, an ant’s alarm pheromone communicates something like “Danger!” to nest mates because in the deep evolutionary history of ant behavior, past ants who perceived that alarm pheromone acted as if there were something dangerous in the environment. In other words, the historical coupling between an ant’s perception of the alarm pheromone and an ant’s engaging in some defensive behavioral routine worked to increase the fitness of ants such that natural selection preserved the relationship between the pheromone and the defensive behavior. According to a teleosemantic understanding of intentionality, natural selection crafts living systems to perform fitness-optimizing functions, and it is that historical process of function formation that underlies intentionality in biology.
When applied to the intentional-talk of ant researchers, however, Millikan’s teleosemantic account of intentionality doesn’t quite fit. Consider the lineage of experimental work on insect navigation that eventually led myrmecologists Matthias Wittlinger, Rüdiger Wehner, and Harald Wolf to demonstrate that the desert ant Cataglyphis measures distance travelled via some kind of odometer or step counter.2 In reviewing that lineage of experimental work, myrmecologist Bernhard Ronacher frames the basic problem of desert ant navigation in the following way: “Obviously, the ants need two kinds of information to determine the home vector. They must combine information about the actual direction of the path – relative to a reference system – with information about the distance travelled in a certain direction.” The first thing to notice here is that Ronacher is employing intentional concepts to frame the problem. Somehow, ants are gathering information about distance and direction of travel. Under a teleosemantic reading, Ronacher’s framing of the problem translates to “some aspect of the desert ant has a special relationship to the properties of direction and distance travelled because natural selection acted on the ant’s ancestors to make it so.”
My contention is not that the teleosemantic translation is unreasonable. Surely, ant researchers would agree that natural selection is the ultimate cause of the desert ant’s ability to successfully navigate its environment. Rather, my contention is that intentional concepts are doing more work for ant researchers than a teleosemantic reading suggests. Consider the hypotheses about insect navigation that were in circulation before Wittlinger et al.’s 2006 experiment.3 Researchers had hypothesized that insects measure distance travelled by keeping track of how much energy they have expended;4 they hypothesized that insects measure distance traveled by monitoring how fast objects flowed through their field of vision,5 as we now know bees do,6 and they hypothesized, as Wittlinger et al. later demonstrated, that insects measure distance travelled via some kind of proprioceptor or “muscle memory” associated with walking movement.7 Under a teleosemantic reading, Ronacher’s framing of the problem doesn’t put many constraints on potential hypotheses. It only requires that the process of ant navigation be grounded in natural selection. The hypotheses that have historically been at play, however, show evidence of further constraints.
For example, the energy hypothesis, the optical flow hypothesis, and the proprioceptor hypothesis all seem designed to meet a robusticity constraint. Whatever the details of an insect’s special relationship to distance and direction of travel are, that relationship needs to be robust in the sense that the relationship will hold in a large variety of environments. The energy hypothesis conforms to the robusticity constraint because insects will have to expend energy to move, no matter the environment; the optical flow hypothesis conforms to the robusticity constraint because most environments will have stationary landmarks that insects can see; and the proprioceptor hypothesis conforms to the robusticity constraint because insects need to move to navigate, no matter the environment. In ant behavior research—and perhaps animal behavior research in general—intentional concepts may imply features about the living system under examination, features like robusticity, that can be cashed out in natural, non-intentional language. By analyzing the knowledge-gathering practices of scientists in addition to the concepts those scientists use to describe their work, philosophers can take an empirically grounded approach to the puzzle of intentionality.
Indeed, ants are an ideal species for such a project. First, unlike molecules of DNA, ants exist at the scale of human experience. That’s important because traditionally, philosophers have been concerned with understanding intentionality in the context of human behavior. Insights about intentionality gleaned through the intentional-talk surrounding DNA is less likely to translate to the realm of human activity than insights gleaned through the intentional-talk surrounding a fellow social organism, the ant. Second, although contemporary eusocial insect researchers sometimes posit the existence of mental representations in eusocial insects, like cognitive maps,8 the field is generally skeptical of such abstract mental phenomena and moves to operationalize such claims at the neuronal level.9 The field’s preference for and ability to produce fine grained physiological descriptions over abstract cognitive descriptions is enormously useful because it provides philosophers with language that is fully naturalized. Contrast that with behavioral research on higher mammals, like dolphins or elephants, that has a harder time producing mechanistic cause-and-effect explanations for reasons involving the complexity of higher mammals and their amenability to laboratory conditions. Ants exist in a sweet spot on the continuum of biological entities. They are human-like enough for us to believe we can learn something about ourselves from them, but they are simple enough to be a tractable system for detailed experimentation.
Some might argue that the empirically-grounded approach to understanding intentionality I have advocated here is bound to return an overly broad account of intentionality. That’s because in the past, when philosophers have attempted to ground intentionality in the cause-and-effect language of systems thinking as opposed to the deep history of natural selection, they often face reductio ad absurdum rejoinders from other philosophers arguing that their cause-and-effect systems account of intentionality grants intentionality to absurdly simple systems, such as the fuel governor on a steam engine.10 In the face of such worries, it is important to remember that there is as yet no satisfactory account of intentionality for any aspect of the world, human affairs included. When philosophers argue that an account of intentionality is too broadly inclusive, they are arguing from their intuitions about how intentional relationships must be. But historically, philosophers have had trouble getting those intuitions to hang together in a way that makes sense of the world. To move forward, philosophers should give more weight to the way scientists use intentional concepts to gather knowledge about purportedly intentional systems, systems like the ant.
1Millikan (1984).
2Wittlinger et al. (2006, 2007).
3Wittlinger et al. (2006).
4Heran and Wanke (1952), Heran (1956), von Frisch (1965)
5Ronacher and Wehner (1995)
6Srinivasan (1996, 1997).
7Pieron (1904), Turner (1907)
8Gould (1986), Cheeseman et al. (2014), Morrison (2014)
9Wehner & Menzel (1990), Dyer (1991), Cruse & Wehner (2011), Cheung et al. (2014)
10See Bechtel (1998) for an account of intentionality that grants intentionality to mechanisms like the fuel governor. See Ramsey (2007) and van Gelder (1995) for arguments that the fuel governor does not exhibit intentionality.
Works Cited
Bechtel, W. (1998). Representations and cognitive explanations: Assessing the dynamicist’s challenge in cognitive science. Cognitive Science, 22 (3), 295-318.
Cheeseman, J. F., Millar, C. D., Greggers, U., Lehmann, K., Pawley, M. D., Gallistel, C. R., … & Menzel, R. (2014). Way-finding in displaced clock-shifted bees proves bees use a cognitive map. Proceedings of the National Academy of Sciences, 111 (24), 8949-8954.
Cheung, A., Collett, M., Collett, T. S., Dewar, A., Dyer, F., Graham, P., … & Webb, B. (2014). Still no convincing evidence for cognitive map use by honeybees. Proceedings of the National Academy of Sciences, 111 (42), E4396-E4397.
Cruse, H., & Wehner, R. (2011). No need for a cognitive map: decentralized memory for insect navigation. PLoS computational biology, 7 (3), e1002009.
Dyer, F. C. (1991). Bees acquire route-based memories but not cognitive maps in a familiar landscape. Animal Behaviour, 41 (2), 239-246.
Gould, J. L. (1986). The locale map of honey bees: do insects have cognitive maps? Science, 232, 861-864.
Heran, H. (1956). Ein Beitrag zur Frage nach der Wahrnehmungsgrundlage der Entfernungsweisung der Bienen. Z. Vergl. Physiol. 38, 168-218.
Heran, H. and Wanke, L. (1952). Beobachtungen über die Entfernungsweisung der Bienen. Z. Vergl. Physiol. 34, 383-393.
Millikan, R. G. (1984). Language, thought, and other biological categories: New foundations for realism. MIT press.
Morrison, Jessica. (2014). Bees build mental maps to get home. Nature News & Comment. https://www.nature.com/news/bees-build-mental-maps-to-get-home-1.15333
Pieron, H. (1904). Du role du sens musculaire dans l’orientation de quelques espèces de fourmis. Bull. Inst. gén. psychol, 4, 168-186.
Ramsey, W. M. (2007). Representation reconsidered. Cambridge University Press.
Ronacher, B., Wehner, R. (1995) Desert ants Cataglyphis–fortis use self-induced optic flow to measure distances traveled. J Comp Physiol A, 177:21–27.
Srinivasan, M., Zhang, S., Lehrer, M., & Collett, T. (1996). Honeybee navigation en route to the goal: visual flight control and odometry. Journal of experimental Biology, 199: 237–244.
Srinivasan, M., Zhang, S., Bidwell N.J. (1997). Visually mediated odometry in honeybees. Journal of experimental Biology, 200: 2513–2522.
Turner, C. H. (1907). Du rôle du sens musculaire dans l’orientation de quelques especes de fourmis. Psychological Bulletin, 4 (9), 296-297.
Wehner, R., & Menzel, R. (1990). Do insects have cognitive maps? Annual review of neuroscience, 13 (1), 403-414.
Wittlinger, M., Wehner, R., & Wolf, H. (2006). The ant odometer: stepping on stilts and stumps. science, 312 (5782), 1965-1967.
Wittlinger, M., Wehner, R., & Wolf, H. (2007). The desert ant odometer: a stride integrator that accounts for stride length and walking speed. Journal of experimental Biology, 210 (2), 198-207.
van Gelder, T. (1995). What might cognition be, if not computation? The Journal of
Philosophy, 92 (7), 345-381.
von Frisch, K. (1965). Tanzsprache und Orientierung der Bienen. Berlin: Springer.
Kelle Dhein is a PhD candidate in the Biology & Society program at Arizona State University. He is currently looking at how animal behavior researchers use intentional concepts to understand living systems. His interests lie at the intersection of philosophy, biology, cognitive science, and anything involving the word “information”. Kelle has been fond of ants since childhood.
The Daily Ant 