Hoffman's Conscious Realism:
A Critical Review

2. Fitness Beats Truth (FBT) Theorem

2.1 Modelling Natural Selection

Book cover: The Mystery of Consciousness by John R. Searle

Hoffman's form of metaphysical idealism, Conscious Realism, relies on an empirical-mathematical theory he and his collaborators term the Fitness Beats Truth (FBT) Theorem. Hoffman [2018: 1] summarizes this theorem like this:

... analysis of perceptual evolution using evolutionary game theory reveals that veridical perceptions are generically driven to extinction by equally complex non-veridical perceptions that are tuned to the relevant fitness functions. Veridical perceptions are not, in general, favored by natural selection.

In their paper illustrating their mathematical modelling, Hoffman and his collaborators reject the dominant view among evolutionary biologists by insisting that 'attempting to estimate the "true" state of the world corresponding to a given a sensory state, confers no evolutionary benefit whatsoever' [Prakasha et al 2017: 24].

I think that Hoffman and his team may have overstated their case here in a couple of ways. Their FBT modelling does not show all that they say it shows. It may show that in the long run fitness beats truth most of the time, but evolution is an ongoing process. So, in the short run, with constant genetic variation resulting from spontaneous mutations and the random pairing of alleles during sexual reproduction, some members of a species will sense truth to at least some extent.

Secondly, even in those cases in which fitness is maximized through selection, there remains some mapping to external reality, even if it is not homomorphic. Otherwise, to use Hoffman and his team's example [Prakasha et al 2017: 9] of organisms competing for a water resource, the organism using the 'Fitness-only' strategy won't get the water it needs to survive. Shermer makes exactly this same point when he writes, 'Finally, why present this problem as an either-or choice between fitness and truth?' [Shermer 2015]. Other writers have also picked up this same point, including Cohen [2015] and Vlerick [2014: 66f] summing up this more nuanced view of natural selection processes.

In fact, in an interview with Frohlich [2019] Hoffman seems to concede this point that our perceptions are truth-tracking to some extent. In replying to the question of why it is that we see the Milky Way when that perception has no fitness payoff, Hoffman says: 'So the idea will be that evolution has shaped us with a very simplified interface that's been shaped mostly [emphasis added] to report the stuff that's going to keep us alive'. Hoffman goes on to say that even so, we misjudge the relative distances to distant mountains, stars and the moon because these distant objects all have the same low 'fitness payoff' for us. So, according to Hoffman, our perceptual systems are 'representing' objective facts about fitness cost and 'report' information to us about objective reality to the extent needed to keep us alive.

More particularly, from the perspective of the science of evolutionary biology, another objection concerns how the FBT Theorem deals with the phenomenon in nature of mimicry. As Dickinson puts the objection:

Hoffman would argue we see an icon that represents a snake, not a snake. But then why do nonpoisonous snakes evolve colorings to match poisonous ones? If there is no objective reality to mimic, why would mimicry prove a useful adaptation, and why would the interfaces of multiple species be fooled by such tricks?

[Dickinson 2019]

Shermer [2015] makes the same objection. I'm inclined to think that Hoffman can answer this kind of objection within his framework without much effort. Could he not respond that what the nonpoisonous snake mimics is not the poisonous colourings of a real snake, but the attributes of the perception of the predator when that predator 'perceives' a poisonous snake? An analogy here is how a character in a virtual reality game can take on the appearance of another avatar in the game when such mimicry is to that character's advantage in the game.

Book cover: Mind and Cosmos: Why the Materialist Neo-Darwinian Conception of Nature Is Almost Certainly False by Thomas Nagel

Mausfeld [2015] advances a more general objection to Hoffman's approach. He finds serious problems with Hoffman's assumption that the notions of 'objective reality' and 'truth' have a role to play in biological theories of perception. Hoffman [2018: 11] responds to this objection by distinguishing between 'proximate' questions about internal mechanisms and 'ultimate' questions about the evolutionary development of those mechanisms. I find Hoffman's answer here a satisfactory response to Mausfeld's objection. Mausfeld also advances a number of conceptual objections to Hoffman's FBT Theorem and takes issue with Hoffman's use of selective constraints in the development of perceptual systems. I think these objections remain unaddressed.

Another objection is that in their mathematical simulations, Hoffman et al's methodology rigs the game in favour of the strict interface strategy. Using Monte Carlo simulations, Hoffman et al [2015a: 1486] try to show that natural selection favours a strict interface strategy over naïve realist and critical realist strategies (shown in their Fig. 1 and Fig. 2 reproduced below). However, they are only able to demonstrate that conclusion using the contrived parameters they fed into their simulation. Hoffman and his team had set each player to perceive only four colours across the whole range of actual resource quantities spanning zero to 100. For the critical realist player, they perceive the colour blue when encountering actual resource quantities between 75 and 100. This methodological choice splits the resource quantities with the maximum payoffs across two perceived colours (yellow and green), increasing the realist players' cognitive load compared with their strict interface competitors.

Now note how if Hoffman and his team had set the perception of blue for both realist strategists to coincide with a resource quantity between, say, 40 and 60, then those players would be as finely tuned as the players using the strict interface strategy. Hoffman et al choosing to break the perceived colours for the realist strategy at precisely the mid-point of the Gaussian payoff curve was an entirely arbitrary choice that just happened to suit Hoffman et al's FBT hypothesis. And that handicap for the realist strategies results from their methodologically arbitrary choice in carving up the resource map with an even number of colours (i.e., 'where the perceptions of each player are limited to just four colors' [2015a: 1486]). If Hoffman and his team had chosen an odd number of colours to input into their mathematical simulation, such as 3, 5, or 7, a single colour would have ranged over the resource quantities with the highest payoffs for both the realist strategies and the strict interface strategy. In Hoffman et al's game, the strict interface strategy was guaranteed a win because they had tied one hand behind the back of the realist strategists.

Hoffman's diagrams for plotting payoff vs resource quantity for critical realist and interface strategies

Martínez [2019] mounts another serious objection showing how Hoffman et al's mathematical modelling is overly simplistic. Here, he provides a number of empirical counterexamples to Hoffman's thesis that fitness always beats truth. He shows that Hoffman's simple model is accurate in only those cases in which an organism receives information from one source. Where an organism makes judgments based on two independent sources of information to ascertain utility, sending and receiving truthful signals promotes fitness. In the next section, I will explore how Hoffman et al's oversimplification creates even more conceptual problems for their FBT theory.

Copyright © 2020, 2022

You will be interested in

Share This

  • twitter
  • facebook
  • linkedin
  • googleplus
  • gmail
  • delicious
  • reddit
  • digg
  • newsvine
  • posterous
  • friendfeed
  • googlebookmarks
  • yahoobookmarks
  • yahoobuzz
  • orkut
  • stumbleupon
  • diigo
  • mixx
  • technorati
  • netvibes
  • myspace
  • slashdot
  • blogger
  • tumblr
  • email
Short URL:
PDF Download Hoffman's Conscious Realism: A Critical Review

Download this essay