Helen De Cruz
School of History, Philosophy and Culture
Oxford Brookes University
Abstract: We have seemings as a result of the ordinary workings of our cognitive faculties (ordinary seemings), and seemings as the result of long-standing deliberate training and practice (skilled seemings). Do these kinds of seemings confer justification in the same way? I argue that, in spite of their similar phenomenology, ordinary and skilled seemings have distinct developmental origins and neurological underpinnings, and that these differences matter for the justification of beliefs formed on the basis of these seemings. I identify three key areas where skilled and ordinary seemings differ: cognitive penetrability, metaphysical structure, and social practice.
Download Helen’s paper in PDF.
Keynote Presentation by Helen De Cruz
Human observers frequently rely on seemings that they only acquire after extensive practice. Birders discriminate species of birds on the basis of subtle cues such as size, shape, color, and habitat. Radiologists can diagnose a patient on the basis of minute differences in grey shade on an x-ray. Even with the naked eye, gemologists can tell apart gemstones that look identical to neophytes through visual properties such as hue, saturation, and luster. These examples involve perceptual seemings, but skills also provide us with intellectual seemings, such as hunches and intuitions. Mathematicians pondering a new theorem may get a sense that it is provable and how it could be proved, before the details of any proof are worked out (Thurston, 2006). A car mechanic, upon looking at a car engine, can get a sense of what is wrong. An epistemologist may form a seeming about whether a given case exemplifies knowledge. My paper will examine whether practitioners of these skills can place trust in seemings like these.
There is no definition of seemings that epistemologists universally agree upon. Some authors (e.g., Hanna, 2011) regard seemings as a kind of non-inferential beliefs, but there are obvious counterexamples. It seems to me that the strawberries of Akiyoshi Kitaoka are red, even when I know they are not. A popular view among epistemologists is the experience view (see e.g., Tucker, 2010), according to which seemings constitute a sui generis experience, which has a distinctive phenomenology; they give the distinct feeling that a proposition is true. Huemer (2001) characterizes this feeling as “forcefulness” and Tucker (2010) calls it “assertiveness”. Given that this characterization avoids some of the problems of other accounts of seemings (see e.g., Moretti, 2015, for an overview), I will adopt it here.
Roughly, for S to have a skilled seeming that p in domain D is for S to have an experience that p is true, as a result of her expertise gained in D, following some stimulus or problem posed in D. For example, a coin collector is presented with a coin, and it seems to her that the coin is an Augustus Denarius. This seeming is the result of her extensive experience with coins of the Roman empire, probably elicited by the shape, size, material and images on the coin. Do skilled seemings confer justification on the beliefs that are formed a result of them? And if they do, is this justification similar to, or different from, a non-trained observer to whom it seems there is a little silver-colored coin with the profile picture of a man on it? Phenomenal conservatism (PC) (see e.g., Huemer, 2001) holds that a subject S’s seeming that p provides S with defeasible, non-inferential justification for believing that p. The (PC) principle does not specify whether the seeming results from ordinary cognitive faculties, or from a highly specialized domain of expertise (e.g., Chudnoff, 2011; Huemer, 2001; Markie, 2005; Pryor, 2000; Tucker, 2010). An example is the following thought experiment by Markie, which juxtaposes non-skilled and skilled seemings as a presumed counterexample to (PC).
Suppose that we are prospecting for gold. You have learned to identify a gold nugget on sight but I have no such knowledge. As the water washes out of my pan, we both look at a pebble, which is in fact a gold nugget. My desire to discover gold makes it seem to me as if the pebble is gold; your learned identification skills make it seem that way to you. According to (PC), the belief that it is gold has prima facie justification for both of us. Yet, certainly, my wishful thinking should not gain my perceptual belief the same positive epistemic status of defeasible justification as your learned identification skills (Markie, 2005, 356).
This is a bad case of ordinary seeming (wishful thinking) compared with a good case of skilled seeming (the training that allows the experienced gold prospector to distinguish gold from, say, pyrite). (PC) is appealing because it provides a unified account of disparate kinds of seeming, including moral, religious, a priori, and perceptual ones (Moretti, in press). In this paper I will show that skilled and ordinary seemings differ in their cognitive structure, and that these differences matter epistemologically.
2. Cognitive foundations of skilled and ordinary seemings
Ordinary seemings are the result of skills that we acquire spontaneously through stable developmental processes. Examples include seemings elicited by ordinary visual perception and intuitions about one’s first language. They do not require explicit teaching or dedicated institutional support, and they are cross-culturally widespread. For example, already in infancy, our visual system makes a spontaneous distinction between movements made by animate and inanimate objects (Kaduk, Elsner, & Reid, 2013). Animate objects tend to be self-propelled, whereas inanimate objects behave according to the rules of our intuitive physics (e.g., they only move when acted upon by an external source).
We do not need to make inferences about whether an object is self-propelled; we perceive it as animate or inanimate. When Hannah sees some purposeful motion in the foliage, it seems to her (thanks to her evolved perceptual ability to detect animate objects) that there is an animal in the foliage. Skilled seemings arise only after extensive practice in a given cultural domain. They include seemings about music, poetry, philosophical and mathematical intuitions, as well as the expert perceptions of birders, car-experts, coin collectors, and the like. They tend to be more restricted in scope than ordinary seemings, typically only available to a subset of experts. To me, a male or female day-old chick look indistinguishable. But to trained chicken sexers, they look distinct. To have such skilled seemings, experts require not only practice, but also teaching and sometimes dedicated institutional support. A birder requires practice and instruction before she can distinguish different kinds of birds, especially if they look and behave similarly. Hannah’s seeming that there is a black-throated green warbler in the tree is a skilled seeming. While it is spontaneous and non-inferential, it required hours of patient observation, learning from experienced birders, and consulting relevant field guides. The distinction between ordinary and skilled seemings is similar to Reid’s (1764 , 171) distinction between original and acquired perceptions: “Our perceptions are of two kinds: some are natural and original, others acquired, and the fruit of experience”.
Ordinary seemings are the result of stable developmental processes and minimal cultural input. Totally disrupting their input leads to atypical development, but this requires unusual circumstances. For example, raising cats in striped environments where the stripes all have the same orientation leads to an over-representation of orientation-selective cells in the visual cortex. This leads the cats to experience their environment in a distorted way (Sengpiel, Stawinski, & Bonhoeffer, 1999). Skilled seemings, by contrast require two key elements: social transmission (especially teaching), and deliberate practice. Deliberate practice is the type of practice learners engage in to become well-versed in a domain of expertise, often centered on improving weaknesses. For example, in chess and music, the number of hours spent in deliberate practice explains about 30% of the variation in expertise (Hambrick et al., 2014). Teaching is required for humans to acquire socially transmitted skills. Studies indicate that pure imitation, for example, tends to be overtly conservative, as exemplified by many instances of overimitation, whereby subjects imitate elements of the skill that are unnecessary (Nielsen, Mushin, Tomaselli, & Whiten, 2014; McGuigan, Makinson, & Whiten, 2011). Teaching, by contrast helps learners to pay selective attention and it also allows them to become innovators, which is an important element of expertise (Fridland, in press), and is probably crucial for acquiring seemings.
The apparent effortlessness with which experts can use their skills in specialized domains can lead to the impression that they possess extraordinary mental capacities for memory and attention to detail, epitomized by fictional characters like Sherlock Holmes. However, an accumulating body of research on expertise, particularly in the domains of music, sports, chess, and writing indicates that these superior capacities are restricted to the domain in which one is an expert, and do not readily transfer to other domains. For example, chess players have a better recall (Chase & Simon, 1973) and can scan a wider visual span of chess positions, but their superior performance does not transfer to other domains of memory or visual attention (Reingold, Charness, Pomplun, & Stampe, 2001). The seemings of experts are thus limited to specific domains of expertise: a gemologist may be able to distinguish gems with similar visual appearance through their luster and hue, but she would not be able to transfer this skill to, say, distinguishing similar-looking orchids.
One reason for this lack of transferability is that expertise relies on specialized neural circuits. Acquiring expertise often involves the redeployment of specialized neuronal systems with older phylogenetic functions in culturally novel domains. For example, the fusiform and occipital face areas underlie face recognition. Car, bird, and dog experts recruit these areas to distinguish between Pontiacs and Oldsmobiles, or between collies and borzois, presumably because like faces, these are members of a same category with visual features that differ in small details. People who are trained to visually discriminate between a novel class of objects that differ in details (greebles) also exhibit the neural signature of face recognition (Gauthier, Skudlarski, Gore, & Anderson, 2000). Nevertheless, visual expertise differs from face recognition in several key respects, particularly in enhanced top-down control. For example, car experts were presented with images of cars and planes, and were asked to only attend to a subset of these stimuli (cars or planes). If expertise were purely automatic and stimulus-driven, one would expect to see similar neural activity regardless of task demands. However, brain areas associated with car expertise were dramatically diminished when car experts had to focus on the planes (where they had no expertise in) (Harel, 2016).
A key element of skilled seemings is control. Cognitive scientists (e.g., Dreyfus & Dreyfus, 1986) used to think about skills as routines. However, studies of sports and music performance have indicated that people are sensitive to context and have control over their skill, unlike over purely routine tasks (Fridland, 2014; Christensen, Sutton, & McIlwain, 2016). Through deliberate practice, brain connectivity is altered in a way that is different from pure routine-formation. fMRI studies show that when novices perform a skilled task a network consistently involved in attentional control, including prefrontal, medial frontal (such as anterior cingulate), posterior parietal, occipito-temporal, and cerebellar areas is recruited (Chein & Schneider, 2005). This network helps the novice to focus on doing tasks in the right order, how to position the body correctly, and to anticipate the next step. Over time, as novices learn the skill, the activity in this network reduces. For instance, as adolescents become better at solving algebraic equations, there is a decrease in activity in their prefrontal cortex, showing they require less executive control to complete the task. Additionally, they also exhibit a decrease in the left parietal region which is normally involved in spatial processing and number, indicating they need to rely less on this as they become more fluent in algebra (Qin et al., 2004). Concurrently, there is an increase over time in activity in modular areas of motor and perceptual processing. The areas vary with the skill that is being acquired. While there is a distinct novice network, there no single expertise network. What distinguishes expertise is changes in activity in domain-specific areas. Neurally speaking, “activity in regions related to domain-specific knowledge distinguishes experts from novices” (Debarnot, Sperduti, Di Rienzo, & Guillot, 2014, 1).
Thus, although skilled seemings and ordinary seemings both rely on modular areas of perceptual and motor processing, skilled seemings are more subject to control than ordinary seemings. This may account for the greater role of background information in skilled seemings compared to ordinary seemings. In particular, high-level cognitive processes play a role in determining the nature of the situation. Once the situation is appraised, the largely automatized processes are adjusted accordingly. For example, a jeweler may use visual appraisal of properties such as luster and brilliance (these are the automatized components of the skilled practice), but before this identification can start, she will inquire about the history of the gemstone which will influence her perceptual appraisals. If she learns that the ring with the stone was bought in a flea market, she will be on the lookout for signs that the ring might be imitation gold, or that the jewel might be fake, whereas if the ring was the client’s grandmother’s heirloom, she would be looking for cues that the stone could be antique paste, an inexpensive substitute for opals, diamonds, and other gems, that was popular in the late 19th to early 20th century.
3 Justification of skilled seemings
Do skilled seemings provide us with prima facie justified beliefs? In this section, I will look at three ways in which skilled seemings are disanalogous to ordinary seemings: cognitive penetrability, metaphysical structure, and social practice. In section 4, I examine whether skilled seemings that p can provide justification for beliefs that p.
3.1 Cognitive penetrability
Skilled and ordinary seemings share a phenomenological feeling of spontaneity and effortlessness. Although both are subserved by largely automatic cognitive processes which, as we have seen, operate at a modular level, there are subtle differences. Skilled seemings are more cognitively penetrable than ordinary seemings. Cognitive penetrability means that the seemings are responsive to the agent’s mental belief states. As Pylyshyn puts it If a system is cognitively penetrable then the function it computes is sensitive, in a semantically coherent way, to the organism’s goals and beliefs, that is it can be altered in a way that bears some logical relation to what a person knows (Pylyshyn, 1999, 343). Ordinary seemings tend to be cognitively impenetrable. The philosophical motivation for not equating seemings with belief states is precisely that seemings don’t always cohere with our beliefs, for example, I know that the strawberries in Akiyoshi Kitaoka’s optical illusion don’t contain any red pixels and yet the picture seems to show red strawberries. But skilled seemings are responsive to my background knowledge. For example, art connoisseurs’ perception of artworks is influenced by their background knowledge of the identity of its creator. A famous example is the Vermeer forger Han van Meegeren, whose works were hailed as masterpieces by the Dutch artist. In particular, his Christ and the Disciples at Emmaus elicited widespread, albeit not universal, praise from art critics. As Robson (2014) points out, after van Meegeren himself showed that the work was a fake, the visual perception of the work by experts shifted markedly. Whereas the colors used to be described as magnificent, splendid, and in perfect harmony, they were later regarded as insipid and lifeless.
What explains the cognitive penetrability of skilled seemings? This is a question of continued debate. According to Fridland (2015, online first), selective attention is key. Experts initiate intentional action, in a way that is sensitive to the condition at hand, and then deploy the selective attention automatically. Once the jeweller knows the ring is my grandmother’s heirloom, she will initiate a series of perceptual processes tuned to late 19th-early 20th century jewelry (e.g., the presence of antique paste), whereas the knowledge that it was bought at a flea market will initiate a different set of attentional processes (e.g., looking for cues that the ring is gold-plated). The grandmaster, looking at a board with an endgame position, will use background knowledge about the level of the players when attending to the position (Reingold et al., 2001). Ordinary seemings are sometimes also cognitively penetrable, e.g., people are typically able to make a gestalt switch between visual illusions such as the duck-rabbit or the old woman-young woman. But skilled seemings are more subject to cognitive penetrability, and as a result, are more under control of the subject.
This greater selective control has several advantages, as it allows the perceiver to attend to features of a given situation that are relevant but it also has disadvantages. First, selective attention sometimes results in inattentional blindness. Inattentional blindness has been demonstrated in naive observers who have to perform an unfamiliar task, namely count the number of ball passes in a video. Many participants did not notice a human in a gorilla suit wandering through the game, as they were concentrated on the task at hand. Expert observers, similarly, are often so absorbed in their expert perception that they miss blatant features that are irrelevant to the task. In one study (Drew, Võ, & Wolfe, 2013), expert radiologists were engaged in a familiar task: to look for lung nodules on five chest CTs. On the final trial, a gorilla outline, about 48 times the size of a typical nodule (the gorilla was about the size of a matchbook) was inserted. In spite of its size, 83% of the radiologists missed the gorilla. Eye-tracking revealed that the majority had in fact directly looked at the gorilla’s location. Presumably, the experts missed the gorilla because they were on the lookout for nodules, which tend to be much smaller.
The theory-ladenness of skilled seemings, exemplified in the altered perception of forgeries, may explain why skilled perceivers do not always accord prima facie evidential value to their perceptual seemings. They frequently cross-check their perceptions by other means, such as different methods (e.g., absolute dating methods to supplement archaeological datings based on style) or by seeking independent validation from other experts. Such methods of checking are far more common for skilled seemings than for ordinary seemings, and are in some cases institutionalized. For example, the micropasts project5 is an online crowd-sourced project that joins archaeologists and amateur enthusiasts in gathering and interpreting archaeological data. One part of the project consists in choosing the best identification (emperor, time period) for specific imperial Roman coins, and asks participants how sure they feel about the identification they provided. Although participants are experts (coin collectors, curators, archaeologists), and have skilled seemings about the coins (“Not Augustus, definitely Tiberius”), there is still a need for validation and cross-checking in a way ordinary perception does not require. For example, no-one would ask if participants are sure the object before them is indeed a coin.
Cross-checking skilled seemings may be partially explained by the high stakes such seemings involve. If the coin expert goes and buys the coin because it seems like a Tiberius in good condition, without consciously going through a checklist, she may be making a buy she will regret. But there are many situations of ordinary perception where the stakes are high, for example, in traffic situations, and where we do not find the need to double-check our seemings. As a result, it is unclear whether (PC) and related principles, such as Pryor’s (2000) dogmatism can be applied to skilled seemings. For Pryor, “when it perceptually seems to you as if p is the case, you have a kind of justification for believing p”, simply by virtue of having an experience as of p (Pryor, 2000, 519). In his view, need not be aware of one’s experience and appeal to facts about it, or appeal to those experiences as evidence. The perceptual experience provides one with an immediate justification, albeit one that is defeasible. The feeling of truth that accompanies some seemings, such as ordinary perception and intuition (assertiveness or forcefulness) thus seems to be weakened in the case of skilled seemings due to its greater control and cognitive penetrability.
The cognitive penetrability of skilled seemings also raises a skeptical worry. What to think of art critics who first praised van Meegeren’s forgeries, and later dismissed them as dull and insipid once the forgery was revealed? What is the evidential value of such seemings (bright, magnificent colors versus dull, insipid ones), if they are so sensitive to background information? Take wine tasters, whose practice has recently come into doubt through a series of experiments that directly probed their expertise. They are influenced by the color of a wine in experiencing its taste. White wines artificially colored to look red elicited olfactory seemings that are commonly associated with red wines in a panel of expert wine tasters. They described the red-colored white wine as “prune”, “chocolate”, “red currant”, words reserved for describing red wine, rather than “honey”, “lemon”, or “butter”, which are commonly used for white wines (Morrot, Brochet, & Dubourdieu, 2001).
Does the presence of faulty background information (the dye in the wine, the art forgery) merely present a set of possible defeaters for skilled seemings, or do they constitute a deeper problem for applying (PC)-type justification strategies to them? Given the importance of background information for skilled seemings, I am inclined to the latter view. I think the prospects of (PC)-style justification strategies for skilled seemings are poor, precisely because they are more influenced by an agent’s beliefs than ordinary seemings are. For an ordinary seeming (which is cognitively impenetrable), the seeming that p can provide justification for my belief that p. If it seems to me there’s a cat on the mat, then that seeming provides justification for my belief that there’s a cat on the mat. But for skilled seemings, my beliefs influence my seemings in such a way that they are not properly independent. Jeanne, the wine expert, drinks the red wine (fortunately not colored red by a malicious cognitive scientist), and it seems to her the wine seems to have an aroma of chocolate, prune and red currant. Jeanne’s seemings are influenced by her beliefs about what red wines smell like. This diminishes the ability of the seemings to justify the beliefs.
3.2 The phenomenal evidence argument
Susanna Schellenberg (2013, 2014) has forwarded a metaphysical argument for the evidential value of perceptual seemings. She argues that they provide evidence due to their metaphysical structure; bad cases of perception, such as illusions and hallucinations, are parasitic on good cases. Take someone who sees a computer (for instance, me as I am writing this text). When I am looking at my computer and forming the belief that there is a computer in front of me, I am using my perceptual capacities that function to pick out particulars, such as whiteness, and shapes, such as rectangles; I also receive haptic input by typing on the keyboard. Through the functional properties of sight and touch, there is a systematic linking between perceptual states and what they are about. In this way, if a subject S is perceptually engaged with her environment, she is in a sensory state that provides phenomenal evidence. As a result, in Schellenberg’s view, even hallucinations provide some evidence (albeit not as much as veridical perception), and it is rational to heed the testimony of our senses. Her view does not require that the perceptual capacities function reliably, and might thus be useful to appraise the evidential value of perceptual skills in the absence of any information about their reliability.
In cases of ordinary perception it is plausible that perceptual capacities pick out properties in our environment. But in some purportedly skilled seemings, there never are any good cases on which the bad cases are dependent. Take as an example soroban (Japanese abacus) calculation. A mathematician skilled with the soroban can read off results of complex calculations by manipulating the beads and looking at the resulting configuration. If he makes a mistake or misreads the result, the output is still dependent on the function of soroban reading (obtaining correct mathematical results) in good cases. By contrast, a trained aura reader “sees” an aura (colorful halo surrounding a human body) as reflecting changes in mood, character, and physical condition. When readingauras, there is never a good case on which the bad cases of aura reading are asymmetrically dependent, because aura reading does not pick out properties in the environment. If we do not know whether skilled seemings pick out any properties in the environment, we cannot claim that there is an asymmetric dependence of good on bad cases. Thus, we cannot use the phenomenal evidence argument to say that skilled seemings confer prima facie justification or that they are some prima facie source of evidence. What we would need is independent evidence that the skilled seeming is indeed picking out what it purports to pick out.
3.3 Dependence on social practice
I have argued in the previous two sections that (PC) cannot be applied to skilled seemings, and that the phenomenal evidence cannot be applied either. And yet it remains plausible that skilled seemings that p provide justification for the belief that p. Skilled seemings are the result of training, teaching and other social practices. I will here examine the justificatory potential of seemings that are heavily dependent on social practice. Kitcher (1995, chapter 7) considered this question for scientific practice. An experienced behavioral biologist observes members of a baboon troop interacting with each other. She sees and hears dominance hierarchies, alliance building, aggression, and submissive behavior. It seems to her the baboon on the right is a dominant female, for example. Her PhD student, who is not yet trained in such observations, just sees a bunch of monkeys moving in a seemingly random fashion. Gradually, he learns to discern the signs that indicate social relationships by the observational skills he acquires from his advisor and other experts.
This intimate dependence of current skilled seemings on past practices can be viewed pessimistically: perhaps the skills are not skills at all, and the heavy theory-dependence of observations could make the interpretations viciously circular. Kitcher, however, resists the pessimistic conclusion that skilled perception in scientific practice would be viciously circular, and offers two reasons for optimism. First, in spite of different perceptual trainings, scientists (and other skilled experts) can come to convergent conclusions. Indeed, Japanese and German primatologists have quite divergent views on cultural transmission in nonhuman primates, which may find their roots in religious legacies in Japan and Germany, with Buddhism stressing continuity between humans and other animals, whereas western culture, with its Christian legacy, emphasizes the uniqueness of human beings (de Waal, 2003). In spite of this, Japanese and German primatologists have convergent results, such as the relative absence of shared attention in chimpanzees (Tomonaga et al., 2004; Herrmann, Call, Hernandez-Lloreda, Hare, & Tomasello, 2007). This remarkable convergence of evidence in the presence of different methodologies and background assumptions should bolster the claim that primatologists in Japan and Germany really are discerning properties of chimpanzee social cognition.
Second, experts can engage in what Kitcher terms “displays of discriminatory virtuosity”: the behavioral biologist can point out subtle differences between monkeys (such as posture and gait), and can predict what will happen next in a social interaction between them (e.g., the subdominant male will back off). Both convergence in the face of different methodologies and a good track record of predictions can help decide if skilled seemings provide one with justified beliefs.
Kitcher’s proposal places the bar for justified beliefs quite high for skilled seemings, much higher than for ordinary seemings. It is not always possible to know whether one’s judgment would conform to the seemings of other experts, and not all forms of skilled seemings allow for predictions, e.g., an art appraiser who assigns paintings to long-dead artists on the basis of stylistic similarities with other works may not always be able to check if these artists indeed created the works, especially not if she is the only expert on an obscure painter.
4 Skilled seemings in an evolutionary and developmental context
Most skilled forms of perception are socially transmitted. When we learn practices that give rise to skilled seemings, could we be led astray by charlatans, who purport to teach us skills they do not really possess? Suppose I would like to learn to distinguish real from fake diamonds (such as cut glass or cubic zirconia), how likely am I to be taught the wrong skills? In game theoretical models of animal communication, the question arises how deceitful communication can be avoided. What prevents communicators from transmitting false information? And, once lies are circulated, what prevents the communication system from breaking down? Acquiring a skill is a lot of work. Given the importance of deliberative practice, several thousand hours are required to master a skill sufficiently to be able to teach it. But once the skill is acquired, demonstrating how to perform it is cheap. In animal signaling, communications that are cheap for a truthful signaler, but hard (or impossible) for a deceitful signaler, can give rise to a reliable communication system that is relatively robust.
As Sterelny (2012) points out, honesty has a byproduct advantage. It is easy for him to present himself credibly as an Australian birder because he really is an Australian birder. Of course, he needed considerable time and effort to learn the requisite birding skills, but once acquired, he does not even need to signal consciously—the evidence for his birding qualities just arises as a byproduct of his daily life. By contrast, someone who wants to present herself as a birder without the requisite skills would need to invest lots of time and effort to be able to pass as such, including at least acquiring some knowledge about birds, and investing in binoculars and a field guide. Not only is knowing-how the norm of skill transmission, as Buckwalter and Turri (2014) have recently argued, it is hard to feign to transmit a skill you do not have, but relatively easy to show and tell a skill you do possess. To take an example from scientific practice, The bluffer’s guide to archaeology (Bahn, 2007) contains real archaeological knowledge (albeit presented in a jocular fashion), which readers would need to bluff their way into archaeology. A more accurate title of the book would have been Acquire some elementary knowledge of archaeological practice so as to be able to pass as an archaeologist among non-archaeologists. Because of the high costs involved in deception when pretending to teach a skill one does not possess, the risk in being deceived when acquiring skilled seemings is low.
While one can discount the risk of deliberate deception in practiced skills, there is still the risk of learning a bogus skill such as aura reading, palmistry, or phrenology. Here, practitioners of such purported skills might not deliberately deceive, but may themselves be deeply mistaken about the nature of their practices. In some cases, the matter can be settled independently by external evidence, such as gathered by the sciences. For example, there is a plausible naturalistic interpretation for why some people see auras: auras may be a particular form of synesthesia, emotion-color synesthesia, whereby a subject associates emotions or people generating these emotions with particular colors (Ward, 2004). Moreover, scientific studies have failed to identify fields of energy that aura-readers purportedly pick up, and under controlled experimental conditions, alleged aura-readers failed to use their skilled seemings to detect people hiding behind screens using their auric emanations (Gissurarson & Gunnarsson, 1997). Taken together, such scientific evidence calls the skilled seemings of aura readers into question. Unfortunately, many skilled seemings cannot be external calibrated in this way. As Cummins (1998) has pointed out, philosophical intuitions do not lend themselves readily to external scientific tests.
One reason for trusting skilled seemings is that they are often part of a broader cultural skillset. For example, to become a proficient chess player, one must be able to see more than mere figurines on a board. One must discern positions, pieces that are overburdened or underdefended, potential forks, and the like. Moreover, to the chess expert, the figurines are tools that perform specific motions, and that have particular effects. When the grandmaster Hou Yifan perceives that a situation is hopeless for White she is applying part of the skillset that makes her an excellent chess player. Given her skills in chess, she can be confident in her seeming that White cannot win this particular game from the current position. Precisely these sorts of seemings are required to be an expert chess player.
Similarly, a trained jeweler can be justified in believing a ring is a rhodolite garnet. Being able to distinguish (semi)precious stones and metals is a skill that requires training and practice, which she presumably received from someone who can signal expertise cheaply (another jeweler). While one can be deceived on the flea-market into believing that a ring is made of 18 carat gold, it is much more difficult to be deceived as a trainee in a jeweler’s workshop who learns to distinguish gold from counterfeit materials. Moreover, being skilled in the art of recognizing real gold from gold plated objects, pyrite, and other superficially similar materials is crucial for jewelers in their daily life. By contrast, the naïve gold digger in Markie’s (2005) example, who really wants to the nugget to be gold has no justification for his belief. It is not easy to distinguish noble metals without training, so there is no prima facie reason for why the belief would be justified, and wishful thinking is in general a bad guide for belief formation.
In this paper, I have examined whether skilled seemings, the seemings one acquires after deliberate practice in a given domain of knowledge, provide us with prima facie justified beliefs. I argue that they in fact do, but that we cannot rely on tactics for the justification of ordinary seemings, such as phenomenal conservatism and the phenomenal evidence argument.
Anderson, M.L. (2007). The massive redeployment hypothesis and the functional topography of the brain. Philosophical Psychology, 20, 143–174.
Atherton, M., Zhuang, J., Bart, W. M., Hu, X., & He, S. (2003). A functional MRI study of high-level cognition: The game of chess. Cognitive Brain Research, 16, 26–31.
Bahn, P. (2007). The bluffer’s guide to archaeology. London: Oval Books.
Brown, S., Martinez, M. J., & Parsons, L. M.(2006). The neural basis of human dance. Cerebral Cortex, 16, 1157–1167.
Buckwalter, W., & Turri, J. (2014). Telling, showing and knowing: A unified theory of pedagogical norms. Analysis, 74, 16–20.
Chase, W., & Simon, H. (1973). Skill in chess. American Scientist, 61, 394–403.
Chein, J. M., & Schneider, W. (2005). Neuroimaging studies of practice-related change: fMRI and meta-analytic evidence of a domain-general control network for learning. Cognitive Brain Research, 25, 607–623.
Christensen, W., Sutton, J., & McIlwain, D.(2016). Cognition in skilled action: Meshed control and the varieties of skill experience. Mind & Language, 31, 37–66.
Chudnoff, E. (2011). The nature of intuitive justification. Philosophical Studies, 153, 313–333.
Copenhaver, R. (2010). Thomas Reid on acquired perception. Pacific Philosophical Quarterly, 91, 285–312.
Cummins, R. (1998). Reflections on reflective equilibrium. In W. Ramsey & M. DePaul (Eds.), The role of intuition in philosophy (pp. 113–127). New York: Rowman & Littlefield.
Debarnot, U., Sperduti, M., Di Rienzo, F., & Guillot, A. (2014). Experts bodies, experts minds: How physical and mental training shape the brain. Frontiers in Human Neuroscience, 8, 1–17.
Dehaene, S., & Cohen, L. (2007). Cultural recycling of cortical maps. Neuron, 56, 384–398.
de Waal, F. (2003). Silent invasion: Imanishi’s primatology and cultural bias in science. Animal Cognition, 6, 293–299.
Drew, T., Võ, M.L.-H., & Wolfe, J.M. (2013). The invisible gorilla strikes again: Sustained inattentional blindness in expert observers. Psychological Science, 24, 1848–1853.
Dreyfus, H.L., & Dreyfus, S.E. (1986). Mind over machine. The power of human intuition and expertise in the era of the computer. New York: Free Press.
Fridland, E. (2014). They’ve lost control: Reflections on skill. Synthese, 191, 2729–2750.
Fridland, E. (2015, online first). Automatically minded. Synthese, 1–27.
Fridland, E. (in press). Do as I say and as I do: Imitation, pedagogy and cumulative culture. Mind and Language.
Gauthier, I., Skudlarski, P., Gore, J. C., & Anderson, A. W. (2000). Expertise for cars and birds recruits brain areas involved in face recognition. Nature Neuroscience, 3, 191–197.
Gissurarson, L. R., & Gunnarsson, A.(1997). An experiment with the alleged human aura. The Journal of the American Society for Psychical Research, 91, 33–49.
Hambrick, D.Z., Oswald, F.L., Altmann, E.M., Meinz, E.J., Gobet, F., & Campitelli, G. (2014). Deliberate practice: Is that all it takes to become an expert? Intelligence, 45, 34–45.
Hanna, N. (2011). Against phenomenal conservatism. Acta Analytica, 26(3), 213–221.
Harel, A. (2016). What is special about expertise? Visual expertise reveals the interactive nature of real-world object recognition. Neuropsychologia, 83, 88–99.
Herrmann, E., Call, J., Hernandez-Lloreda, M.V., Hare, B., & Tomasello, M. (2007). Humans have evolved specialized skills of social cognition: The cultural intelligence hypothesis. Science, 317, 1360–1366.
Huemer, M. (2001). Skepticism and the veil of perception. Lanham: Rowman & Littlefield.
Kaduk, K., Elsner, B., & Reid, V.M. (2013). Discrimination of animate and inanimate motion in 9-month-old infants: an ERP study. Developmental Cognitive Neuroscience, 6, 14–22.
Kitcher, P. (1995). The advancement of science. Science without legend, objectivity without illusions. Oxford: Oxford University Press.
Markie, P. (2005). The mystery of direct perceptual justification. Philosophical Studies, 126, 347–373.
McCauley, R.N. (2011). Why religion is natural and science is not. Oxford: Oxford University Press.
McGuigan, N., Makinson, J., & Whiten, A. (2011). From over-imitation to super-copying: Adults imitate causally irrelevant aspects of tool use with higher fidelity than young children. British Journal of Psychology, 102, 1–18.
Moretti, L. (2015). Phenomenal conservatism. Analysis, 75, 296–309.
Moretti, L. (in press). Phenomenal conservatism and the problem of reflective awareness. Philosophical Quarterly.
Morrot, G., Brochet, F., & Dubourdieu, D. (2001). The color of odors. Brain and Language, 79, 309–320.
Nielsen, M., Mushin, I., Tomaselli, K., & Whiten, A. (2014). Where culture takes hold: “Overimitation” and its flexible deployment in Western, Aboriginal, and Bushmen children. Child Development, 85, 2169–2184.
Pryor, J. (2000). The skeptic and the dogmatist. Noûs, 34, 517–549.
Pylyshyn, Z. (1999). Is vision continuous with cognition? The case for cognitive impenetrability of visual perception. Behavioral and Brain Sciences, 22, 341–365.
Qin, Y., Carter, C. S., Silk, E. M., Stenger, V. A., Fissell, K., Goode, A., & Anderson,
J.R. (2004). The change of the brain activation patterns as children learn algebra equation solving. Proceedings of the National Academy of Sciences USA, 101, 5686–5691.
Reid, T. (1764 ). An inquiry into the human mind, on the principles of common sense. A critical edition (D.R. Brookes, Ed.). Edinburgh: Edinburgh University Press.
Reingold, E. M., Charness, N., Pomplun, M., & Stampe, D. M. (2001). Visual span in expert chess players: Evidence from eye movements. Psychological Science, 12, 48–55.
Robson, J. (2014). A social epistemology of aesthetics: Belief polarization, echo chambers and aesthetic judgement. Synthese, 191, 2513–2528.
Schellenberg, S. (2013). Experience and evidence. Mind, 122, 699–747.
Schellenberg, S. (2014). The epistemic force of perceptual experience. Philosophical Studies, 170, 87–100.
Sengpiel, F., Stawinski, P., & Bonhoeffer, T. (1999). Influence of experience on orientation maps in cat visual cortex. Nature Neuroscience, 2(8), 727–732.
Sterelny, K. (2012). The evolved apprentice. How evolution made humans unique. Cambridge, MA: MIT Presss.
Thurston, W. (2006). On proof and progress in mathematics. In R. Hersh (Ed.), 18 unconventional essays on the nature of mathematics (pp. 37–55). New York: Springer.
Tomonaga, M., Tanaka, M., Matsuzawa, T., Myowa-Yamakoshi, M., Kosugi, D., Mizuno, Y., Okamoto, S., Yamaguchi, M.K., & Bard, K.A. (2004). Development of social cognition in infant chimpanzees (Pan troglodytes): Face recognition, smiling, gaze, and the lack of triadic interactions. Japanese Psychological Research, 46, 227–235.
Tucker, C. (2010). Why open-minded people should endorse dogmatism. Philosophical Perspectives, 24, 529–545.
Ward, J. (2004). Emotionally mediated synaesthesia. Cognitive Neuropsychology, 21, 761–772.
 For a similar distinction, see McCauley’s (2011) maturational and practiced naturalness.
 However, Reid carves the conceptual space up a bit differently from how I do it. He provides examples that involve some form of expertise and thus would fall under the category of skilled seemings, such as a sailor who “sees the burthen, the built, and the distance of a ship at sea, while she is a great way off” (Reid, 1764 , 172). But he also categorized forms of perception that require no deliberate practice or teaching, such as the smell of an apple or an orange, as acquired perception (Reid, 1764 , 171). There is a continued discussion on whether acquired perception is as direct as original perception, and whether there is any inference involved in acquired perception, see Copenhaver (2010) for a defense that acquired perception, even the highly skilled forms (which fall under the category of skilled seemings), are indeed direct.
 The view that culturally novel practices can co-opt the functional properties of phylogenetically older neurological substrates is termed “neural reuse”, “massive redeployment” (Anderson, 2007) or “cultural recycling” (Dehaene & Cohen, 2007).
 Pryor was not sure in how far his account could accommodate skilled perception, because he did not know in how far skilled perception is basic (Pryor, 2000, 539).
 Practitioners of bogus skills often are part of superstitious communities that disregard external, invalidating evidence, or are unaware of it.
 The medial temporal gyrus, which is activated when people see tools such as hammers and saws, is more strongly activated in chess experts than in novices when seeing chess pieces (Atherton, Zhuang, Bart, Hu, & He, 2003).
21 thoughts on “KEYNOTE: Skilled Seemings”
Hi Helen! Cool paper! As a former advisee of Mike Huemer, this is right up my alley.
When I was reading the paper, I found myself thinking about a similarity between ordinary and skilled seemings: they seem to benefit from deliberation, or “cross-checking” (regardless of task difficulty) (Moxley et al 2012). Two interesting things might follow from this. On the one hand, this similarity could be a problem for one of your claims. On the other hand, this similarity could be an opportunity to strengthen your claim about PC. Here is what I mean:
Hi Helen: very interesting talk! I still need to read the paper, but here’s a quick query: I wonder what implications your analysis of skilled seemings has for philosophers’ appeals to intuitions. As I’m sure you know, X-phi studies often indicate that folk-respondents don’t make the same judgments about cases that philosophers have. For example, Tobia et al’s study reported at this conference indicates the folk don’t make the same semantic judgments about Twin Earth that philosophers have. One common reply to studies like these–the “expertise defense” (viz. Williamson, etc.)–seeks to attribute these differences to a kind of philosophical expertise. However, as your example of the aura-readers suggests, a major problem with “skilled seemings” (including philosophical appeals to intuition) is that they can be subject to serious sociological confounds. One fabulous study I recall (in a very different context) found that when an identical batch of songs was presented to different samples of respondents, songs that were initially judged “good” by people in one sample were then also rated as good by other people in that sample (as a kind of snowball effect)–even though, in other samples, a corresponding snowball led to the exact opposite results. Given how susceptible “expert judgment” can be to such confounds, aren’t these just additional reasons to wonder about the expertise defense? Classical philosophical Twin Earth judgments, Gettier judgments, etc., could be examples of expertise…but they could also just be examples of snowball effects where one “expert judgment” leads subsequent “experts” to make similar judgments. Do you think your analysis has implications here?
Hi Marcus – thanks, those are good points (the original paper was longer and had a section on philosophical intuitions, but ultimately it was not possible to do it justice in the paper so I cut it). I am skeptical of the expertise defense. In my 2015 paper, “Where philosophical intuitions come from” (AJP) I argue that philosophers do have expertise when it comes to intuitions, namely to think of cases (thought experiments) that elicit intuitions. But the intuitions they have aren’t better than those of laypeople (plenty of evidence to suggest philosophers are subject to framing effects etc.).
I think there’s two kinds of philosophical intuitions – ones that everyone has, e.g., Gettier intuitions which are cross-culturally quite robust (see Machery’s work), and ones that are specific to philosophers (e.g., swampman intuitions and other exotic scenarios). Now, if I’m right in this paper then we should be more confident about the ordinary seeming philosophical intuitions than we should be about the specialist, skilled philosophical intuitions. This is because there is no clear way in which philosophers can calibrate their intuitions to the external world (unlike, say, the primatologist).
Hi Nick – thank you, I did not know the Moxley et al study, it is very relevant to this paper! I would need to think more about the relationship between cross-checking and (PC). Luca Moretti has a nice paper out about (PC) where he argues an agent can lose her justification provided by (PC) when she reflects on the seemings. Only small children and animals, or situations where this reflection would be unavailable, would still have (PC)-based justification (see here: https://philpapers.org/archive/MORPCA-8.pdf) – Now, while it is true that both skilled and ordinary seemings benefit from cross-checking, we do not generally cross-check in the case of ordinary seemings (we sometimes do, especially if the seeming is ambiguous). As Moretti says, “it is a matter of fact that we seldom engage in the practices that would destroy our seeming-based justification in everyday life, as we customarily don’t reflect on our seemings and don’t speculate about their cause”. But he thinks maybe we should reflect more on our ordinary seemings. I am not convinced that this conclusion is warranted. There are ecological constraints on how much time we can spend reflecting on our seemings (Did I really see a cat on the mat? etc.). If that’s the case then the justification of skilled and ordinary seemings differs because, practically speaking, (PC) is both less relevant to skilled seemings (as we have reflect on them) and less applicable (because of their cognitive penetrability).
This is helpful clarification. Thanks Helen! And thank you for the link to Moretti. That’ll is a timely recommendation for my dissertation research and on my thinking about reflection vs. intuition more generally.
Thanks Helen, for a super interesting paper.
I’m not sure whether you’ve come across it, but Ryle talks about a related phenomenon in the Concept of Mind. I think the discussion is around when he talks about how knowledge of philosophy is important for properly understanding Aristotle.
I just wanted to pick up and push back on your claim that skill-transmission is i) easy, and ii) resistant to deceit. I definitely agree that there are some things that are easy to show other people, where there are features that are easy for the non-skilled to pick up on, and where success is obvious. Maybe there’s a special knack to unlocking the door to my house that requires a wiggle that’s easy to recognise and quickly leads to success. But I worry that not all cases work like that. If the features of a skilled action are sufficiently complex, then skill transmission is not easy: you can’t just show someone how to do a double back somersault on a trampoline — you need to pull the act apart and show them all the different parts, and then how they fit together. Similarly, if the feedback loop between demonstration and success is sufficiently long, we’re definitely vulnerable to deceit with skill transmission. For example, it I wanted to find out how to train to run a really fast 10k, the feedback between getting advice and the actual performance (which might be some months) opens me up to getting misleading advice. I can’t just fall back on looking for people who can succeed themselves and try to find people who run fast 10ks, because quick people might not be good at formulating advice, and good trainers might not be quick themselves. And even someone who has put years into training might just be mistaken about the best way for me to train to run a fast 10k. I guess I’m thinking that there will be similar examples with the transmission of recognitional abilities practiced skills (i.e. real skills, not like aura reading which I take it you’re thinking of as a separate case). For example, trained GPs who have put in significant time to their diagnostic skills might still be mistaken about what various rarer conditions look like, and if they don’t get feedback once a referral, it could be a real threat that they pass on these mistaken seemings to their trainees.
Woah – ended up a bit long. I’m a bit puzzled by how the cases pattern here, and I’d be really interested to hear what you think!
I think you overstated justificatory roles in your examples. If “skilled seemings” rise to the level of justified belief for the purposes of expert inquiry, then why does the gemologist with a seeming bother using a refractometer; the mechanic or the doctor get their hands dirty; the birder have a book; the primatologist even bother running their study? In these examples “seemings” seem like maybe initial sparks guiding productive expert inquiry, but that they probably wouldn’t be relied on too much beyond that, let alone count as justified without further tools, proofs, diagnostics or established procedures in these fields.
Thank you for a very stimulating paper! I found myself somewhat puzzled by the distinction you were drawing between ordinary seemings, which develop spontaneously and benefits from (PC)-style justification, and skilled seemings, which do not because of the centrality of background information in their etiology (it seems to me that you think this is a necessary condition for skilled seemings). I’m wondering about cases where background information is involved in generating the seeming but where (i) this information is not a particular belief but just a sensitivity that has developed “spontaneously” within the expert’s practice of the skilled activity and (ii) the seeming involves a change merely in the salience of certain features of the environment. Say, a mathematician looks at a worksheet submitted by her research assistant and a particular equation stands out to her. She successfully spots the mistake. This ‘seeming’ is certainly colored by background information—even though plausibly, this is done by a process that is reliant primarily on the affective system which imbues the particular line with a ‘glow’, instead of employing processes as fine-tuned as those involved in your example of the jeweler whose perception is “tuned to late 19th-early 20th century jewelry”. More broadly, I’m wondering if the different ways in which skill influences our seemings is relevant to their prima facie justification.
Hi Wesley – yes, it seems that often the skilled perception is just the initial hunch and there’s often this extensive cross-checking. Maybe the exceptions are such cases as wine experts and art connoisseurs (and wine experts turn out to be quite variable in their ratings, so it does not seem to bode well for them), so maybe I need to rephrase the question as: why do skilled perceivers bother to cross-check? Maybe it is because their seemings do not have the same justificatory potential as ordinary seemings. Philosophers are among those experts who continue to rely on seemings (e.g., knowledge intuitions in cases), and I think that is a potentially interesting example as it is difficult for philosophers to cross-check.
Thank you so much – I really like the 10k example. There’s not a very good philosophical understanding yet on how we can transmit knowing-how, and the paper briefly touches on that but I realize more would need to be said. Some people, like Ted Poston recently, argue that you can’t transmit knowing-how through testimony (others, like Cath, think that you can), but their idea about what counts as testimony is quite narrow (as in purely verbal). If you broaden the scope of transmission to have both showing and verbal testimony, I think you’re right that there is scope both for deceit and for suboptimal information transfer (e.g., your trainer may wrongly give you the advice to start really quickly, whereas it’s proven better to start slowly when running 10k). If the skill consists of demonstrations, it is definitely vulnerable to conceit. I am thinking of complex skillsets where it would still be far more costly for the teacher to deceive than to be honest. For instance, in the birding example I would need to come up with fake bird names, and fake cues. It’s definitely easier to be a genuine birder and give the correct information. The skills I have in mind are quite complex, which is why, while not impossible to deceive, it is easier to signal honestly (which might be enough for reasonable skill transmission)
One possible answer is that experts are sensitive to something like the analogue strength of a seeming (“epistemic feelings”), and use this through executive control to decide when to trust the hunch and when to consult external decision aids. The gemologist may trust a strong seeming but consult a refractometer when the seeming is experienced as weak, or evidential standards are very high (a big or certified sale). The birder may trust a strong seeming for a well-known or distinctive species, but not a weak seeming for one that is rarer or has close lookalikes. For domains that can support genuine expertise (i.e. offer quick feedback to diagnose success and causes of failure, etc.), experts do tend to be well-calibrated in their confidence.
Whether this is good enough for high-church epistemology justification depends on one’s view of the standards of justification that ought be applied, especially whether they are internalist or externalist. My view is that the right standards for this kind of thing are a kind of hybrid depending upon the kind of question you are asking, with explanatory questions appealing to internalist criteria like intensional fit that grants agent-level executive control (through epistemic feelings), while justificatory questions should appeal to externalist criteria that depend upon things like ecological validity. But that’s a longer story…
Sorry, to clarify my comment above was directed at Wesley’s question, and several other replies popped up in the interim!
Helen: I agree they are often initial hunches that would not satisfy many experts. I also think it’s interesting philosophy is different.
Cameron: I think you are right experts are sometimes sensitive to the character of seemings and that this could potentially be instructive. I also really like your approach of framing things with respect to specific research questions. But I doubt sufficiency is a matter of “high-church epistemology”. Surely the researcher in Helen’s key example would not rest with mere “seeming” for justifying claims about the nature of baboons.
Helen: I think you’re exactly right — understanding how we gain skills/know-how is super duper complicated and that makes it hard to even think about the possibility of deceit. In addition to the cases of testimony and demonstration you consider (and what exactly is demonstration anyway!), there will be cases where someone picks up a bit of know-how just by watching someone else doing something without the person being watched intending to teach them.
On the birdwatching example, it definitely seems right that it is easier for a skilled birder to be truthful than it is for an non-skilled person to make stuff up. But what about how easy it would be for a semi-skilled birder (or even a fully skilled birder) to make stuff up? For (say) a super skilled birder, wouldn’t it be pretty easy to make up claims about what species a bird is by drawing on their background information and knowledge? Or at least, it seems like it wouldn’t be so hard as to make it unfeasible.
To be clear, I’m replying to Wes again here…hoping comment threading will work this time!
The baboon-dominance example is an interesting one. For most purposes, the expert primatologist would almost certainly trust seemings about something like dominance (and even appeal to seemings/expertise if justifying things to a graduate student). This might change if they entered into a different context where the standards for justification were raised–for example, publishing a paper about dominance in a high-profile scientific journal, where they might be expected to do an inter-coder reliability check, or report data on the results of conflicts. But they might also be expected there to post videos of the conflicts and share other data; but I’m not sure there’s any special problem with the seemings in this kind of case, the idea is just that the stakes are so high that we want others to be able to scrutinize every part of the knowledge-making enterprize, all the way down to the raw perceptual input.
When I talk about “high-church epistemology”, I just mean cases where really explicit, discursive reasons-talk is really appropriate–the good old sorts of explicit, coherence-based norms of classical rationality. I think these standards are too high for most decision-making at the individual and even everyday social level (where more ecological standards are more appropriate), but I hope they continue being applied to publication in Nature or PNAS!
Cameron: Yes, I also hope those standards remain in place exceeding seemings for the central way experts and professionals proliferate inquiry. I also agree this can be more demanding then everyday life, which is to cut against the original point of the paper. Incidentally, I’m a little unclear what dominance “seemings” as a clsss actually amount to in the primatologist case, over and above applying knowledge and reasoning in a pretty straightforward way.
Wes-Well, to get back to what Helen says about convergence in the paper, maybe it’s helpful to think closer about things like inter-coder reliability tests. The way this would typically be done is that a primary author would recruit some of the other authors or members of the lab–often grad students–to redundantly code some subset of the recorded interactions. Sometimes you’ll get an explicit coding sheet for this–what to look for in the interactions–but sometimes you’ll just get a bit of training by the researcher. Once you think the graduate student is reliable enough at identifying the relevant properties (i.e. dominance) in the judgment of the expert, multiple researchers independently code some subset of interactions and see how much agreement there is in the judgments.
Now, is this really a matter of appealing to some separate, more rigorous, non-seeming-based source of justification? Or is it just a way of formalizing the kind of novice-to-expert calibration and convergence that Helen talks about in the paper?
(It remains a very interesting question if/why philosophy as a discipline should be judged differently…)
To answer your specific question about why “seemings” might be treated differently as a class: I think seemings are involved in a lot of straightforward knowledge-application and reasoning, so maybe we don’t disagree. I still think it remains useful to talk about cases where seemings were involved and when they were not because seeming-talk serves to mark cases where specific grounds for judgment are opaque.
Hi Helen. Fascinating paper! I wonder whether there’s a further disanalogy between skilled and ordinary seemings. In skilled seemings, agents often seem to have some grip on why things seem that way to them. When asked why the jewel seems to be rhodolite garnet a jeweler might be able to cite things like its luster and brilliance. By contrast, when it seems to the prospector that the nuggets are gold there’s not really a lot they can say about why it seems this way to them. It just does!
What I’m suggesting isn’t the same as saying ‘experts have justifications for their beliefs and novices don’t’. I’m allowing that in both cases we’re talking about seemings which have a status quite different to that of beliefs. Even if it doesn’t make sense to justify our seemings, we can try to identify properties that corroborate what our seeming reports. I’m also not appealing to control here. Whether or not an agent can control her seemings by attending to luster or to context is distinct from whether she can cite luster and context to corroborate her seemings.
So perhaps one of the differences between skilled and ordinary seemings is that access that subjects have to the process responsible for that seeming. In the ordinary cases there’s something brute about the appearance whereas in the expert case there’s some insight into how the seeming arose. This is simply the inverse of cognitive penetrability: instead of high-level cognition influencing the processes responsible for the skilled-seeming, that process is disclosing itself to high-level cognition.
I think this is right – it is a result, I think, of the training required for the skilled seemings. Much of that training is in an explicit, verbal format, and gives some info about the causal processes that are at work (e.g., in the case of the jeweler you get trained with the relevant talk about lustre, brilliance etc.). I will need to think more about how it relates to cognitive penetrability and control – I agree with you that it is a distinct property, albeit I think closely related to these other properties.
My question may be too late, but here goes. I’m wondering about evolution and skilled seemings. It looks as though with neural reuse we can get functions that are not selected for by evolution. This in turn puts pressure on the thesis that has functions connected to evolutionary selection. That might not seem a problem unless skilled seemings are more
My iPad decided I had written enough, so just stopped with my previous entry. The concluding comment should be “an integral part of lives.” I think I’d best stop here.
Comments are closed.