Jump to content
Objectivism Online Forum

Rational Recurrence: Cognitive Capacities and Conclusion (Part 3; End)


Recommended Posts

Cognitive Capacities



To be sure, there are perceptual-like capacities such as imagination. They are not capacities which function to single out and discriminate particulars; these capacities by nature are not concerned with concrete particulars in the world, but with particulars that have already been discriminated and singled out. Their function is to entitle attitudes by deriving information directly from PE. They reveal something about the world and therefore provide a basis to take some attitude towards the world, even if not presentational like perception. The perceptual-like nature is inessential to the role they serve for attitudes, though. What counts for attitudes in general is that information is extracted from PE. Even capacities that are far from perceptual-like, such as intuition, emotion, and reason, all use PE to produce attitudes. They may provide attitudes besides rational ones.


No matter how non-perceptual capacities are characterized, they at least function to provide entitled attitudes, so serve as a transition from Vladmir’s perception of a flying object to developing his narrower, rational attitude to shoot the object down. As an example of how non-perceptual capacities work: mice navigate mazes with cognitive maps in mind. A radical behaviorist would want to say a mouse navigates by stimulus and response – in a perceptual way without use of imagination or other abstraction abilities. But a more robust explanation that explains their behavior is with some additional mechanism like a cognitive map [10]. While the science isn’t proof of the philosophy, it does illustrate my principle. Whatever the mechanism is precisely, the key idea is that perceptual capacities provide perceptual evidence which is then used by another capacity to produce an attitude. The attitude – in non-conceptual terms for the mouse itself - may be one where a mouse does not regard a left turn as a means to the cheese, or where it would regard its perceptual evidence as suggesting there is nothing strange about the maze, e.g. there aren’t rotating platforms.


Vladmir would act analogously after he has been presented perceptual evidence by his perceptual capacities. First, his perceptual capacities present him with definite perceptual evidence that he is seeing an object in the sky which is moving in a circle. From that evidence, he develops an attitude. The attitude may be one of truthiness - that in his gut, the object just has to be a dangerous weapon used by the Ukranians. Another possible attitude is that he feels an existential need to assert radical freedom in the chaos of war and fire his weapon just to show that he can. Still another attitude is the rational one, where reason is used to determine the attitude that his weapon should be fired. If existentialist-Vladmir repeated the situation, he just as easily could refrain from firing the weapon, or take any variety of actions for eternity during recurrence. While I’d argue that he’d take the same attitude for eternity in spite of his belief in radical freedom, the very nature of a rational attitude means the real Vladmir used a definite method to reach a definite answer that isn’t dependent on approximations.  


            Emotional capacities use PE to establish emotion – it’s why Vladmir felt concerned in response to seeing the flying object. Capacities of intuition produce a “feeling” in Vladmir to suspect a real threat without any reasoning. Whatever capacities exist, the point is that they generate attitudes, parallel to how a PC creates a SS. I classify an attitude as a non-perceptual state with a “likeness” to it, impacting the behavior of an agent because it makes an agent aware of information for action. Such division is not arbitrary; it is based on introspection, where emotions, intuitions, and sensory states are distinguishable. Not that an agent will always properly distinguish them, but that it is possible. The issue I’m concerned with, however, is doxastic attitudes – attitudes pertaining to explicit propositions of beliefs, opinion, or knowledge. More specifically for Vladmir, his doxastic attitude is rational, i.e. is justified and meets a normative standard. He will be able to tell that he is in a DA, and be able to tell if his DA is rational, all in terms of phenomenal character that attitudes provide. The story so far about a rational DA follows the same pattern as a veridical SS. The essential remaining issue, though, is how a DA can be distinguished in the first place from other phenomenal states, and how it is possible to determine which DAs are rational. The necessary mechanism is a cognitive capacity, which by nature works to generate rational DAs.


Cognitive capacity refers to a type of mental capacity which is non-perceptual. Although cognitive typically refers to anything related to the mind, as in cognitive science, I want to distinguish the means of cognition from unqualified mental functioning. The word ‘cognitive’ in this context denotes a narrower concept than usual in order to specify seeking the truth; the CC is the sole means of cognition, the sole means of producing knowledge if knowledge is to ever be factive. It evaluates evidence of many kinds, perceptual or not, with some type of reasoning standard, as opposed to habit, heuristic, or instinctual impulse [11]. In this sense, the phenomenal character of a DA is equivalent to the feeling of making a judgment, the consequence of a CC completing its job to evaluate evidence. Certainly, cognitive psychology could demonstrate more details, or even if a CC is broader than I imagine. The point is that it’s what makes Vladmir identifiably rational. To be identifiably rational, he requires the ability to employ a CC without unnoticed intervention of other capacities, and requires that his rational DA is grounded in the determinacy of PCs through a connection with CCs resulting in a singular rational DA.


            A CC is an a priori mechanism, meaning it is able to operate without any PE, or anything underlying. Using the same principle of structure-form relation used to discuss PCs, it produces the correct form all without requiring any PE, and by implication, requiring no SS. Specifically, a CC is a priori because it doesn’t require having any experiences of the world at any point. For instance, a writing an argument over the merits of democracy requires no SS in the whole process, or even for deriving underlying premises and concepts for the argument, but the process of writing an argument exhibits the content a CC is meant to generate – rational DAs.


Because a CC is an a priori mechanism, neither the form nor the structure of cognitive capacities can malfunction. That is, the capacity’s structure will always produce the right form for a DA. Damage to the brain or body cannot make the capacity malfunction. Accordingly, agents are only able to lose or lack a cognitive capacity entirely. Consider that if a CC were working to a reduced degree of competency, it still produces a judgment as to the status of the agent’s perceptual evidence. The judgment isn’t necessarily warranted, nor does it need to be, because at a minimum, the CC only needs to produce some DA in order to be evaluated as rational or not – a lack of warrant does not point to a malfunctioning structure. In contrast, a broken PC fails to produce an SS which can be evaluated as PE, so it fails on the level of form at the very least. But if not a failure of form, capacities in general can in principle fail on the level of structure. The only sensible way that I see to say a CC’s structure malfunctioned, though, is to say that its DA lacks warrant, a state which still was produced by a CC and evaluated as unwarranted. However, this malfunction is parasitic upon a CC’s structure already having standards of rationality applied to it. As long as a CC exists, it is in a position to provide evidence.


            The theory of mind mechanism is cognitive capacity that is used to understand the mental state of others. All CCs work like the ToMM, meaning it has all essential features. If the ToMM did not, it would be something else entirely: it might be a non-cognitive mechanism; it might only function to produce emotional content; it might be a PC. But none can produce a DA. Research suggests [12] that those with autism have a diminished ability to understand the mental states of others, due to lacking a ToMM or the ToMM malfunctioning. If the reason for difficulty is lack of a ToMM, then there is no CC to malfunction. If the reason for difficulty is a malfunctioning ToMM structure, then I would suspect a deformed DA, analogous to cases of kaleidoscopic vision or drug trips when a PC is malfunctioning. But it’s not as though the DA is deranged. The supposed malfunctioning is an example of the CC containing strange content, yet content that can be grasped quite reliably and easily to anyone with a ToMM, autistic or not. Kaleidoscopic vision can be described in words, except as an SS, it isn’t in a form that is commensurable with the typical form produced by a PC – it’s the wrong “shape”, a shape which the PC was never built for. Compare this to the odd content of an autistic individual’s DA, where the odd form is commensurable with the typical form produced by a CC. Ask an autistic child where Sally will look next, and if the answer is “where the doll is now”, then it can easily be evaluated and compared to an answer “where it was before”. The consistency of form makes the DA the right shape for the CC’s function. 


Application of rationality is where a malfunction can occur in a CC, not in the structure itself. The CC is able to work without regard for rationality, but the CC will fail to operate according to its purpose without rationality. That’s why, despite how a CC is always able to entitle its DA, it’s the application of rationality as a method that determines if a CC does entitle its DA. For a PC, the equivalent of rationality is measurement. The key difference is that rationality must be agent-directed in entirety, while measurement doesn’t require the same deliberate effort. After all, my ability to discriminate a particular tree in front of me via measurement doesn’t take any deliberation that a written argument requires to reach a sound conclusion. The standards that make up rationality are not obvious to say the least, but a CC cannot entitle a DA as rational without a method to operate on the content the DA consists of. Rationality doesn’t come for free by just having a DA - in comparison to a SS which is inherently discriminated, without conceptual intervention, as baseless or not by the phenomenal character alone. So, even with the seeming infallibility of a CC if its structure and form can never go wrong, and always being in a position to entitle, the resulting DA may still be baseless. Until rationality is applied to a DA, a CC cannot have sought the truth, as it would still be unknown if the DA were linked to reality.


Interestingly, with the purpose of rationality to ground a DA in reality, it is possible to specify agent-relative rational decision making while still holding strict normative standards of how rationality needs to operate – not just any standard of rationality will accomplish grounding. Evidence established by rationality will be interested for the agent, as such grounding orients his actions towards reality based on his own PE. This accomplishes helping an agent find appropriate actions in his environment that are oriented towards reality, and preventing him from failing to reach his goals just in case he has the wrong facts. More simply, if one’s goal is to build an airplane, it’s necessary to get the facts right, or else the airplane would be pure imagination, luck, or a wreck at the bottom of a ravine. Whether or not building an airplane is itself a rational goal can be debated, but rationality would go through the same process. I don’t take the rules of rationality to be anything aside from rules of reasoning and logic, but this characterization I’ve presented is premised on the precision of PCs, which creates a fixed point where rationality must begin [13].


I’m supposing there isn’t any more involved in generating a rational DA, so if rationality is applied successfully, then the DA is necessarily rational, and entitled as evidence.  


Fuzzy Attitudes?


Vladmir will only ever have one rationally permissible DA. His CC won’t operate in any other way during recurrence. A singular permissible set of PE, as apprehended by a SS, has already been shown for Vladmir, but how can his DA become equally definite?


Even with the successful application of rationality to his PE and any earlier formed evidence, the right attitude isn’t narrowed down to one in all cases. Even if PE is definite, and rationality is applied properly, that doesn’t mean there is enough evidence for a rational DA. When there’s (supposedly) only partial evidence to justify a DA, it’s possible to say the DA is incompletely rational, with a probabilistic weight of 75/25 for shooting/not shooting. If a DA could not be any finer, as a rational man, Vladmir would be perfectly right to shoot 75% of the time during recurrence, and not shoot 25% of the time. This would be as rational as Vladmir can be, and I’d have to fall back on some non-capacity view where rationality can only come from reliable capacities. A Bayesian could easily endorse this view.


The putative partial evidence doesn’t need to be a problem. Although Vladmir doesn’t have the evidence he would like (e.g. that the flying object is carrying arms, has no innocents on board, doesn’t contain a crucial head of state to negotiate a treaty, etc), he isn’t entitled to any additional rational DAs. After all, I don’t say that if I don’t have all possible knowledge of commercial flight contingencies, I am not able to hold a rational attitude towards going on a flight. I only lack evidence for specific details. This means that evidence is all or nothing for rational attitudes, as there is still a singular DA that my CC points to, e.g. boarding the flight is right to do. Vladmir may still hold a rational attitude towards probabilities regarding what could happen or how pro-value an action is, yet his CC and application of rationality will still use that probability as evidence for exactly one attitude. If the weight of shooting is 75%, that weight can then be used to say “I may be only 75% sure that the flying object is an enemy drone, but this evidence I’ve created from PE and rational rules to definitively determine Bayesian weights, it is all evidence used to achieve my wartime survival. Altogether, only shooting the drone can be rational, since it is a superior option for my goal“. The partial evidence is in fact complete for his rational DA, and the only attitude which the evidence can entitle, as long as a rational DA is an attitude of interestedness.


A Bayesian-esque view is the strongest one able to hold a notion of indeterminacy while still maintaining definite rational standards – indeterminacy could only lead to probabilities or weights. To be more succinct and to consider more than just Bayesian views, what counts is that insufficient evidence for something Vladmir wants to know motivates a rational DA; refraining from judgment can be a rational attitude. Vladmir wants to know that he isn’t about to shoot down a plane of Ukranian civilians, but he can’t know that fact any more than he can know the color shoes his friend from college is wearing today. The rational attitude would be that he will not act with regard to what he doesn’t have total evidence to entitle.


Condensed Argument


Eternal recurrence as a thought experiment shows that employed capacities do not provide different evidence in identical situations. If a capacity succeeds in providing evidence it was made for, it has acted according to its nature. A failure of the capacity means evidence is lacking for any rational attitude later on, meaning there is no evidence to speak of, only warrant at best.


When a PC operates, its standards of measurement with the particulars it discriminates in the given context provides a single set of PE and a corresponding SS to recognize it. So, holding identical PE implies an identical PC, further implying that the measurement standards for discrimination of particulars are identical. Even more, for particulars to be discriminated identically, the context and particulars need to be identical. Any deviation makes PE different. Vladmir has no way to change his PE in recurrence if he wants a rational DA.


            The same pattern applies when CCs operate. Its standards of rationality with the attitudes it discriminates as rational in the given context provide a single set of evidence and a corresponding DA to recognize it. So, holding identical evidence implies an identical CC, further implying that the rationality standards for discriminating rational attitudes are identical. For attitudes to be discriminated identically, the context and attitudes are also identical. Any deviation makes evidence different. A CC cannot operate in a grounded (and rational) way without PE, so a failure of a PC causes a CC to fail, ensuring that capacities must operate as an integrated whole for rational DAs. Vladmir has no way to change his evidence in recurrence if he wants a rational DA.


In the end, if two people hold identical evidence, they will both have only one rationally permissible attitude towards a given situation. Thinking of Vladmir’s recurrence makes it easier to see this principle.


But does acting rationally in such a definite and prescribed way diminish the value to be had and created in life? That’s the deeper question Nietzsche wanted to answer with his hypothetical – how can anyone embrace their choices if they’re not even permitted to choose otherwise?


PC - perceptual capacity
SS - sensory state
PE - perceptual evidence

CC - cognitive capacity
DA - doxastic attitude

Edited by Eiuol
Link to comment
Share on other sites


10. c.f. Randy Gallistel, Organization of Learning. Animal learning, even for insects, shows complex patterns of behavior without any conceptual abilities or reasoning.

11. c.f. Jonathan Haidt, 2001. If reasoning can only be used for different ends than cognition such as a post-hoc rationalization, then CCs would be entirely different and not about knowledge. Haidt uses this term to describe how an intuitive belief exists when encountering an idea, then reasoning is applied only to rationalize an intuition as true. I don’t deny the phenomena, but I would not say reasoning is only meant for rationalizations.

12. Alan Leslie, “ ‘Theory of Mind’ as a Mechanism of Selective Attention”, in The New Cognitive Neurosciences, second edition, Gazzaniga, ed. I am especially referring to the Sally-Ann style experiments mentioned in Leslie’s chapter.

13. c.f. Michael Tittelbaum, “Rationality's Fixed Point (Or: In Defense of Right Reason)”. I adopt Tittelbaum’s terminology of fixed point because his fixed point asserts that some beliefs about rationality cannot be considered permissible, that there are facts about rationality which one rationally cannot disagree with or be rationally be mistaken on. If one was mistaken, that belief could not have been rational. I don’t quite support all the arguments he used to demonstrate his thesis, but his view is one similar to mine: 1) there are definite constraints on the requirements of rationality, 2) rationality may not have self-undermining rules, and 3) “if an agent has correctly determined what attitudes her situation requires, rationality forbids changing those attitudes in the face of apparent evidence that she's made that determination incorrectly”. Vladmir is that agent, so rationality requires he holds the same attitudes in recurrence.

Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.


  • Recently Browsing   0 members

    • No registered users viewing this page.
  • Create New...