In Search of Dennett’s Free-Floating Rationales

I’ve decided to take a closer look at Dennett’s notion of free-floating rationale. It strikes me as being an unhelpful reification, but explaining just why that is has turned out to be a tricky matter. First I’ll look at a passage from a recent article, “The Evolution of Reasons” [1], and then go back three decades to a major exposition of the intentional stance as applied to animal behavior [2]. I’ll conclude with some hints about metaphysics.

On the whole I’m inclined to think of free-floating rationale as a poor solution to a deep problem. It’s not clear to me what a good solution would be, though I’ve got some suggestions as to how that might go.

Evolving Reasons

Dennett opens his inquiry by distinguishing between “a process narrative that explains the phenomenon without saying it is for anything” and an account that provides “a reason–a proper telic reason” (p. 50). The former is what he calls a how come? account and the latter is a what for? account. After reminding us of Aristotle’s somewhat similar four causes Dennett gets down to it: “Evolution by natural selection starts with how come and arrives at what for. We start with a lifeless world in which there are lots of causes but no reasons, no purposes at all.” (p. 50).

Those free-floating rationales are a particular kind of what for. He introduces the term on page 54:

So there were reasons before there were reason representers. The reasons tracked by evolution I have called “free-floating rationales” (1983, 1995, and elseswhere), a term that has apparently jangled the nerves of more than a few thinkers, who suspect I am conjuring up ghosts of some sort. Free-floating rationales are no more ghostly or problematic than numbers or centers of gravity. There were nine planets before people invented ways of articulating arithmetic, and asteroids had centers of gravity before there were physicists to dream up the idea and calculate with it. I am not relenting; instead, I am hoping here to calm their fears and convince them that we should all be happy to speak of the reasons uncovered by evolution before they were ever expressed or represented by human investigators or any other minds.

That is, just as there is no mystery about the relationship between numbers and planets, or between centers of gravity and asteroids, so there is no mystery about the relationship between free-floating rationales and X.

What sorts of things can we substitute for X? That’s what’s tricky. It turns out those things aren’t physically connected objects. Those things are patterns of interaction among physically connected objects.

Before taking a look at those patterns (in the next section), let’s consider another passage from this article (p. 54):

Natural selection is thus an automatic reason finder that “discovers,” “endorses,” and “focuses” reasons over many generations. The scare quotes are to remind us that natural selection doesn’t have a mind, doesn’t itself have reasons, but is nevertheless competent to perform this “task” of design refinement. This is competence without comprehension.

That’s where Dennett is going, “competence without comprehension” – a recent mantra of his.

It is characteristic of Dennett’s intentional stance that it authorizes the use of intentional language, such as “discovers,” “endorses,” and “focuses”. That’s what it’s for, to allow the use of such language in situations where it comes naturally and easily. What’s not clear to me is whether or not one is supposed to treat it as a heuristic device that leads to non-intentional accounts. Clearly intentional talk about “selfish” genes is to be cashed out in non-intentional talk, and that would seem to be the case with natural selection in general.

But it is one thing to talk about cashing out intentional talk in a more suitable explanatory lingo. It’s something else to actually do so. Dennett’s been talking about free-floating rationales for decades, but hasn’t yet, so far as I know, proposed a way of getting rid of that bit of intentional talk.

The rationale of “free-floating rationale”

Back in 1983 Dennett wrote a major article expounding his notion of intentional systems in Brain and Behavioral Science, a journal which, as many of you know, publishes commentary along with the original article [1]. The major aim of the article is to defend the idea that it is permissible to treat animals as intentional beings, where “intention” is meant in the philosophical sense, which is somewhat broader than but inclusive of the ordinary sense. Once he’s gotten going he introduces one of his prime examples of free-floating rationale, though he doesn’t employ the term until after he’s give the example (p. 350):

A phenomenon that will nicely illustrate the connection I wish to draw is “distraction display,” the well-known behavior, found in many very widely separated species of ground-nesting birds, of feigning a broken wing to lure a predator that approaches the nest away from its helpless inhabitants (Simmons 1952; Skutch 1976). This seems to be deception on the bird’s part, and of course it is commonly called just that. Its point is to fool the predator. Now if the behavior is really deceptive, if the bird is a real deceiver, then it must have a highly sophisticated representation of the situation.

That is, it is one thing for us as external observers to assert that the bird is deceiving a predator, but is that what the bird itself intends. That is to say, is the bird “a real deceiver”? Dennett goes on:

The rationale of such deception is quite elaborate, and adopting R. Dawkins’s (1976) useful expository tactic of inventing “soliloquies,” we can imagine the bird’s soliloquy:

I’m a low-nesting bird, whose chicks are not protectable against a predator who discovers them. This approaching predator can be expected soon to discover them unless I distract it; it could be distracted by its desire to catch and eat me, but only if it thought there was a reasonable chance of its actually catching me (it’s no dummy); it would contract just that belief if I gave it evidence that I couldn’t fly anymore; I could do that by feigning a broken wing, etc.

Talk about sophistication! It is unlikely in the extreme that any feathered “deceiver” is an intentional system of this intelligence. A more realistic soliloquy for any bird would probably be more along the lines of: “Here comes a predator; all of a sudden I feel this tremendous urge to do that silly broken-wing dance. I wonder why?” (Yes, I know, it would be wildly romantic to suppose such a bird would be up to such a metalevel wondering about its sudden urge.)

First note that, following a tactic he has from Dawkins, Dennett imagines what a bird would have to be thinking if it intended to deceive. I’ll return to that in a later post. He then asserts, reasonably enough in my opinion, that the bird does no such thing.

And yet Dennett does not want to give up the notion the bird is being deceptive. He reasons (p. 351):

But suppose it turned out that the killjoy interpretation was closest to the truth; the bird has a dumb tropism of sorts and that’s all. Would we thereupon discard the label “deception” for the behavior? Yes and no. We would no longer credit the individual bird with the rationale of deception, but that rationale won’t just go away. It is too obvious that the raison d’être of this instinctual behavior is its deceptive power. That’s why it evolved. If we want to know why this strange dance came to be provokable on just these occasions, its power to deceive predators will have to be distilled from all the myriad of other facts, known and unknown and unknowable, in the long ancestry of the species. But who appreciated this power, who recognized this rationale, if not the bird or its individual ancestors? Who else but Mother Nature herself? That is to say: nobody. Evolution by natural selection “chose” this design for this “reason.”

Why doesn’t he end that third sentence after “deception”? Rocks don’t have to know they’re hard in order to be hard; plants don’t have to know they’re transducing energy from the sun in order to do so. And now it would seem that birds don’t have to know they’re being deceptive in order to deceive.

What’s problematic about that? Well, if we saw a human behaving in that way, we’d say they’re being deceptive and if we asked them what they’re up to they’d tell us: “I’m being deceptive.” As ordinarily understood, the concept of deception implies that the agent consciously intends to deceive. Since birds and other creatures cannot do that, we’ve got an ontological mismatch, if you will, between the agent and the action we’re attributing to that agent. Such a mismatch is ordinarily known as a category mistake.

Why don’t we let it go? Why don’t we simply modify the ordinary sense of “to deceive” to cover those other cases, of which, apparently, there are many? I’m quite quite happy to do that, to introduce a technical sense of “deceive” that differs from the ordinary sense.

But Dennett wants to know where that behavior came from. It didn’t come from the bird’s conscious deliberation, but it must have come from somewhere. So where? “Mother Nature”, Dennett says. He knows that’s a figure of speech, as do we; and he knows that we know. That’s why he invoked it. He’s arguing that evolution is in effect an intentional system, albeit a dumb one, like a thermostat.

Now he’s ready to introduce his term (p. 351):

Is it unwise to speak this way? I call this the problem of free-floating rationales. We start, sometimes, with the hypothesis that we can assign a certain rationale to (the “mind” of) some individual creature, and then we learn better; the creature is too stupid to harbor it. We do not necessarily discard the rationale; if it is no coincidence that the “smart” behavior occurred, we pass the rationale from the individual to the evolving genotype. This tactic is obvious if we think of other, nonbehavioral examples of deception. No one has ever supposed that individual moths and butterflies with eye spots on their wings figured out the bright idea of camouflage paint and acted on it. Yet the deceptive rationale is there all the same, and to say it is there is to say that there is a domain within which it is predictive and, hence, explanatory. […] We may fail to notice this just because of the obviousness of what we can predict: For example, in a community with bats but not birds for predators we don’t expect moths with eye spots (for as any rational deceiver knows, visual sleight-of-hand is wasted on the blind and myopic).

The transmission of the rationale from the individual back to the genotype is of course an old trick. [And so forth…]

Except for the term “free-floating rationale” I’ve got no problem with any of that. It’s the term itself that bothers me and it bothers me because it is a useless reification, perhaps even a category mistake.


By now it should be clear however that Dennett doesn’t introduce the term to tell us about the psycho-neural mechanisms of bird behavior; it’s not about what the bird thinks. The phenomenon that interests Dennett is a property of that situation, which is a specific pattern of interaction between adult bird, nestlings, predator, and various background entities.

The rationale (about the bird’s behavior) that is so obvious to Dennett as an external observer cannot be posited of the bird itself. That reason is an idea in Dennett’s mind (and mine and yours). But bird’s do not, so far as we know, have minds capable of such ideas.

Dennett does not want to be driven to the idealist notion that the rational exists only as an idea in his mind. He in effect says, “Well, it’s somewhere out there in the world, it’s free-floating.” What’s wrong with saying, “No, the rationale is a concept in our conceptual system and, in this case, it corresponds to a pattern that exists in the world”?

Because that’s what I’m going to say. But I’m not, at this point, going to try to tease out the implications. I note that I’ve said a bit about patterns [3] and so, I’ve just discovered, has Dennett [4].

Our discussions are somewhat different. Interestingly enough, his discussion is motivated by a desire to justification talk of intentions, though he doesn’t bring up free-floating rationales. I’ve said nothing about intentions, but do talk about ecological niches [5]. What I’m wondering is whether or not to think of patterns as epistemological objects or as ontological objects. I’m thinking the latter is what I want, but that’s another argument for another day.


[1] Daniel Dennett. The Evolution of Reasons. In Contemporary Philosophical Naturalism and Its Implications, eds. B. Bashour and H. D. Muller. Routledge 2014: pp. 47-62.

[2] Daniel Dennett. Intentional systems in cognitive ethology: The “Panglossian paradigm” defended. Brain and Behavioral Science 6, 1983: 343-390.

[3] William Benzon. Beyond Quantification: Digital Criticism and the Search for Patterns. Working Paper, February 2015.

[4] Daniel Dennett. Real Patterns. Journal of Philosophy. LXXXVIII, 27-51, January 1991.

[5] William Benzon. Pattern as Term of Art. July 5, 2014. Patterns as Epistemological Objects. July 8, 2014. Both posted at New Savanna and included in 4.

1 thought on “In Search of Dennett’s Free-Floating Rationales”

  1. Well, I think that you’re right that “free floating rationale” is an unnecessary reification. Perhaps the big question is whether being unnecessary is a sufficient reason to abstain. If you take a skeptical position on meaning, then you drive out the meaning, don’t you? As an engineer and former AI guy, I tend to favor Dennett’s positions. If we talk about design, we need rationale. Conventionally rationale needs an agent, but none is found here. Rather than saying “oops, that’s inconsistent”, he brazens it out. Free floating rationale is a property of evolutionary design. There are good reasons for wanting to apply design thinking to evolution. See

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.