Intuitions and Perceptions: An Evaluation of Evidential Weight in Epistemology

By Ahmed Elsayyad, Johns Hopkins University

This paper was presented at Prometheus’ 2014 Mid-Atlantic Philosophy Conference.

I. Background

Intuitions are quick and ready insights without any apparent rational thought. There has been debate among the philosophical community on whether intuitions can be used as reliable evidence in answering questions in epistemology. Studies have shown that intuitions can vary by factors such ethnicity and gender. If intuitions can vary by such factors, can we still say intuitions can be used as reliable evidence for philosophical arguments? Some argue that the psychological sources of intuition render it too error prone be used as sound evidence for philosophical discourse. On the other hand, defenders of intuition have conceded that intuition is error prone but they argue that other forms of trustworthy epistemic evidence such as perceptions* are similarly fallible. This raises the question of whether or not we can ever rule out intuitions as having evidential weight while giving evidential weight to perceptions.

II. Introduction

In the following sections, I will argue that intuitions, unlike perceptions, do not carry evidential weight in philosophical arguments. In Section III, I will outline and critique Jennifer Nagel’s defense of intuitions as reliable evidence for epistemic arguments that she lays out in her article, “Intuitions and Experiments: A Defense of the Case Method in Epistemology.” In Section IV, I will introduce and reinforce Jonathan M. Weinberg’s criteria for trustworthy evidence in his “How to Challenge Intuitions Empirically Without Risking Skepticism.” In Section V, I will apply Weinberg’s criteria to perceptions in order to support the argument that perceptions can be used as sound evidence. On the other hand, in Section V, I will illustrate why intuitions ultimately do not meet Weinberg’s criteria. I will then conclude in section VI by showing how we can keep perceptions as evidence in our arguments and still rule out intuitions as evidence.

III. Nagel’s Defense for Intuitions

Jennifer Nagel defends intuitions as reliable evidence for epistemic arguments in her article, “Intuitions and Experiments: A Defense of the Case Method in Epistemology.” Assuming that three premises (N1, N2 and N3) are true, Nagel’s argues (N4) to reach the conclusion that intuition can be used as sound evidence for epistemic arguments. Her argument is as follows:

(N1) Perception is fast, automatic and domain specific

(N2) Perception is reliable

(N3) Reliable judgments are sound evidence

(N4) Intuition is like perception

(Conclusion) Therefore intuition is sound evidence.

In this section, I will suggest that premises N1, N2, N3 are true, N4 is not entirely true and her conclusion is therefore unjustified.

N1. Perception is fast, automatic and domain specific

Scientific literature supports Nagel’s first assumed premise (N1) that perception is fast, automatic and domain specific. Nagel claims that we have “automatic calculations of what an observed agent can see” (Nagel 502). That is, we do not consciously calculate in order to perceive. In his book, Cognition and Perception: How Do Psychology and Neural Science Inform Philosophy, Athanassios Raftopoulos discusses how our neural pathways work in a way independent of our conscious control. When we see something, for instance, we do not consciously tell our brain to direct the stimulus detected by our eyes to be processed by our brain. Raftopoulos states that sensory information emerges from an immediate interaction between the brain and environment and is “domain-specific, fast, automatic, mandatory and independent of conscious control.” Had sensory information not been fast and automatic, there would be a significant delay in processing external stimuli. Thus, we can see that Nagel’s first assumption (N1) coincides with the current scientific understanding of perception.

N2. Perception is reliable & N3. Reliable judgments are sound evidence

Nagel states experimentalists would not dispute the “epistemic legitimacy” of using perceptual judgments as sound evidence (Nagel 496). This assumption is supported by the current accepted scientific method. In Our Knowledge of the External World as a Field for Scientific Method in Philosophy, Bertrand Russell affirms how “the facts of sense-perception are those which we most obviously and certainly come to know.” To ignore the facts of sense perception, then, would be taking on a radical form of skepticism. We can, then, accept Nagel’s premises N2 and N3.

N4. Intuition is like perception

Nagel argues that intuition is fast, automatic, domain-specific and is fallible in much of the same way perception is. Nagel states, “the mechanics of intuitive judgment reveals some deep similarities between epistemic intuition and perception” (Nagel 496).

Nagel uses scholars Hugo Mercier and Dan Sperber’s distinction between intuitions and reflective judgments to highlight intuition’s fast and automatic nature. Nagel argues that because of the “capacity limitations on conscious attention, reflective thinking is sequential in character; where intuitive judgment can integrate large amounts of information very rapidly” (Nagel 499).Nagel claims that the inherent nature of intuitive judgment requires intuitions to be fast and automatic.

Lastly, Nagel acknowledges while intuitions are fallible, they are still fallible in the same way perceptions are. Nagel begins by stating perceptions are fallible but we know when our perceptions are susceptible to illusions. For example, Nagel suggests that one understands the contrasting effects of judgment of color. For example which inner square is darker in the images below (Figure A: extracted from her article)? How confident are you that the inner square you chosen is darker?

Figure A:

Nagel implies that our confidence level for such a question would be lower than a question that did not include the outside squares. (Figure B)




Figure B:





Nagel presupposes that we understand that the context of the outside square may influence our perception of the inside square. Nagel brings up Alexander Swain’s statement that, “We are aware of the great majority of the circumstances under which perceptual judgments are likely to be unreliable” (Nagel 517). Nagel believes that this contextual influence can be also applied to intuitions. In the same way, we are able, to a degree, to determine the validity of our intuitions based on our confidence of them.

Nagel concludes that because premises N1, N2, N3 and N4 are true, intuitions are reliable evidence. In the following sections, I will argue why premise N4 is not entirely true for while there are many similarities to perceptions and intuitions, there are fundamental differences that prevent intuitions from carrying evidential weight in arguments.

IV. Weinberg’s Criteria for Trustworthy Evidence

I will now introduce and reinforce Jonathan M. Weinberg criteria for trustworthy evidence in his “How to Challenge Intuitions Empirically Without Risking Skepticism.” Because we know that perception is fallible, but we still believe perception to carry evidential weight, we cannot suggest that fallibility alone is an epistemic problem. But does that mean intuitions also carry evidential weight? Weinberg alleviates this problem by suggesting that there are multiple ways to be fallible. Weinberg states that what “philosophical intuition[s] are guilty of, but which our other standard sources of evidence are not, is unmitigated fallibility—a fallibility uncompensated by a decent capacity for detecting and correcting the errors that it entails”(Weinberg 323). With unmitigated fallibility as the new condition, Weinberg argues that it’s the ability to correct errors that allows perceptions to carry evidential weight rather than intuitions. Weinberg then suggests that any putative source of evidence that is hopeless ought not be trusted (Weinberg 327). He defines hopeless as anything that does not meet one of the following mitigating criteria.

(W1) Weinberg describes external corroboration as the ability to validate a source of evidence’s results using other methods. We can understand that the necessity of corroboration is to know about when and how these sources of evidence sometimes become skewed. Without this criteria being met, we cannot verify which data is correct and which is faulty.

(W2) Internal coherence is described as the “agreement both within and across subjects” of a particular source of evidence (Weinberg 330). This criterion allows us to reproduce results and ensure the results were not an odd occurrence such as due to chance.

(W3) Weinberg asserts detectability of margins is the ability to know when the device is more prone to give poor results. This allows us to be more cautious in certain contexts of obtaining results. This criterion may prevent us from obtaining results that are predisposed to be inaccurate.

(W4) Lastly, Weinberg illustrates theoretical illumination as the understanding of how a device works when it does and why it fails to work when it doesn’t. By understanding how something works, we can be more confident in deciding whether a source of evidence yielded an accurate result or merely faulted.

V. Weinberg’s Criteria Applied to Sensory Perceptions

I will now apply Weinberg’s Criteria to perceptions in order to highlight how the capacity for mitigated fallibilities allows perceptions to carry evidential weight.

(W1) External corroboration

Sensory perception consists of the five senses: sight, hearing, taste, smell and touch. Since, we have different senses, we are able to corroborate a sense by other senses. For example, let’s suppose a man sees a melted bar of chocolate on a plate. Before he eats the bar, the man wants to make sure the bar is in fact chocolate. The man may corroborate his sense of sight with his sense of smell. The man then smells the chocolate bar and verifies his first conclusion. Had the man smelled something else, he may have reconsidered his first conclusion resulting from his sense of sight. We can then say that because the man was able to mitigate a potentially fallible device (sight), with another sense that perception meets the external corroboration criteria.

(W2) Internal coherence

The very fact that empirical evidence, knowledge acquired by observation or experimentation, is one of the most trusted forms of evidence indicates the internal coherence within perception. Let us suppose that the aforementioned man looking at the chocolate bar was unsure of whether or not his eyes were deceiving him. He may also be doubtful of his other senses. The man can ask another individual whether they can see the melted chocolate. Another man looks and sees in fact it appears to be melted chocolate. The original man is satisfied and eats the melted chocolate. If the original man’s senses were at fault, the original man is able to check with other people’s senses. Because the original man is able to mitigate potential fallibilities with his sense of sight with other people, we can say perception has the check of internal coherence.

(W3) Detectability of margins

We are able to determine when our perceptions are more susceptible to fault. Lets go back to the man looking at the melted chocolate. Lets say he was inside a room that was dimly lit. Because the man is not able to see the melted chocolate clearly, he may not want to eat the chocolate. He knows when not to trust his senses completely. Because we are able to detect when to trust our perceptions, we are able to say that perception meets the criteria detectability of margins.

(W4) Theoretical illumination

Lastly, physicians and scientists comprehend how our senses work when they are functional and when our senses do not work when they are not functional. Let us suppose that the original man was stung by a rare poisonous bee earlier that morning and is now looking at the melted chocolate. The effects of the poison are now manifesting by blocking the original man’s vision so he is unable to see the melted chocolate. He later goes to an optometrist who tells the man that some sort of toxin caused his blurred vision. The doctor is unsure of how the man obtained the toxin but is sure a toxin entered the man’s body. So although scientists may not have a complete understanding of perception, scientists understand theories regarding the mechanisms that allow eyes to function and not function proving that perception meets the theoretical illumination criteria.

Using the sense of sight as an example, I have now illustrated how perception meets all four of Weinberg’s criteria in mitigating fallibilities. This suggests that while perceptions are fallible, they still carry evidential weight because of their mitigating capabilities.

VI. Weinberg’s Criteria Applied to Intuitions

I will now apply Weinberg’s Criteria to intuitions to illustrate how intuitions are hopeless and carry no evidential weight.

(W1) External corroboration

To classify all intuitions as either having external corroboration or not having corroboration would be meaningless. For surely there are many intuitions for which we can corroborate such as mathematics and sciences. For example, if I had asked someone to quickly determine what 35×4 equaled, I can check one’s answer by using already established mathematical concepts. In such cases, where there is another measure of accuracy, I would suggest intuitions do have means of error detection and subsequently have external corroboration. For cases in which intuitions are used to determine epistemic information, there lacks a concrete error detection mechanism. Can we verify people’s thoughts on epistemic norms? Is it possible for what we believe to be true, to be far from the objective truth? If this is even a possibility, we cannot merely establish what everyone believes to be true as absolute truth. One, such as Weinberg, may argue that we can use epistemic norms of the past and present to “speculate counterfactually about what results various sorts of norms might or might not generate for us today” (339). We run into a tautological problem when using the past as an error detection mechanism because we cannot be sure that what human beings, of all times, believe to be true knowledge is true knowledge. Furthermore, this undergoes the assumption that human intuitions across factors such as cultures, genders, age, etc. are similar when we know this is not the case. How can we detect which group’s intuitions are correct and incorrect? In cases where there is no external source of error detection, intuitions fail to have external corroboration.

(W2) Internal coherence

Recent research has shown that intuitions do not have internal coherence, that is, results from intuitions do not have agreement across subjects. Studies have shown that intuitions can vary by factors such ethnicity, socioeconomic status and gender. If intuitions have intrasubject disagreement how can we check and compare the results based on intuitions? This suggests the lack of internal coherence as a mitigating factor in the fallibility of intuitions.

(W3) Detectability of margins

It is uncertain whether we are able to distinguish when our intuitions are correct. Nagel suggests otherwise. Utilizing the results illustrated in Asher Koriat’s Self-Consistency Model (SCM), Nagel argues that individuals have the ability to detect, to an extent, the validity of intuitions. Koriat created a study in which he asked English-speaking participants to match antonym pairs, in random order, in Thai to their English translations. They were then asked to rate their confidence in their choices (1-4 where 1 = wild guess, and 4 = most confident). The results suggest that participants who were confident about their choices had more valid answers: 53.35% of the pairs to which they had assigned confidence level one were correctly matched, while 66.10% of the confidence level four pairs were corrected. Koriat created further studies to test how accuracy was effected when the whole group of students were confident and concluded that “a person’s level of confidence in an intuitive judgment predicts the degree to which that individual will make the same judgment again when presented with the same problem again.” These results suggest that not only do intuitions have domain-specific properties but their validity can also be detected by confidence.

Using these Koriat cases, Nagel argues that we can tell our intuitions are correct with our feeling of confidence. Korait’s results suggest that participants who felt confident about their choices were correct 66.10% of time while those unsure were correct 55.35%. Is a 10% difference in accuracy significant enough to suggest that we are able to detect intuitions based on confidence? I beg to differ, because an individual can be very confident about an incorrect intuition.

In their book, The Invisible Gorilla: And Other Ways Our Intuitions Deceive Us, Christopher Chabris and Daniel Simmons describe how humans are able to believe confidently vivid and remarkable false memories. They bring up the example of Hilary Clinton confidently claiming she was under sniper fire when landing in Bosnia while media coverage revealed her being greeted at a ceremony with no snipers. The authors suggest that this is consistent with their findings of memory distortion. Additionally, the authors conducted a survey on experienced tournament chess players and found that the weaker players significantly overestimated how good they were — players thought they would beat someone with their own skill level more than 2/3 of the time. Stronger chess players did not show as much overconfidence. The authors describe the results of the study as the ‘illusion of confidence principle’ which is that the worst performers are the most overconfident which is “applied to domains ranging from senses of humor to logic abilities.” So confidence cannot be an accurate margin detector for intuitions because we cannot detect whether one has false confidence. Thus, we are unable to detect the margins of fallibility for intuitions.

(W4) Theoretical illumination

The inherent nature of intuitions suggests a lack of understanding of how they work. Hugo Mercier and Dan Sperber describe intuitive judgments as “processes that take place inside individuals without being controlled by them” and the “spontaneous inferences update what we believe without the individual’s attending to what justifies this modification.”Because we have relatively low understanding of how our intuitions work we are unable to know why they work when they do and why they do not work when they do not.

VII. Conclusion

To conclude, Jennifer Nagel argues that because intuitions and perceptions are fast, automatic, domain specific and fallible, and perception is reliable as sound evidence, it follows that intuition is also sound evidence in epistemic arguments. However, I have argued that while both have similar properties, epistemic intuitions, unlike perception, lack Weinberg’s four criteria of error-detection: (W1) External corroboration, (W2) Internal coherence, (W3) Detectability of margins, (W4) Theoretical illumination. Because epistemic intuition lacks these error detection properties, it is impossible to verify if intuitions are reliable or unreliable at any given occurrence. Sensory perception, on the other hand, contains these error-detection mechanisms, which allows us to know when, why, and how perceptual judgments are fallible. Therefore, because of error detection capabilities, perceptions carry evidential weight while intuitions do not.

Works Cited

i Weinberg, J. S., Nichols, S. and Stich, S. (2001). “Normativity and Epistemic Intuitions.” Philosophical Topics,

29(1), 429–460.

ii Buckwalter, W. and Stich, S. (2011). “Gender and the Philosophy Club.” The Philosophers’ Magazine (52), 60–65.

ii Buckwalter, W. and Stich, S. (2011). “Gender and the Philosophy Club.” The Philosophers’ Magazine (52), 60–65.

iii Cummins, Robert. 1998. “Reflections on Reflective Equilibrium.” In Rethinking Intuition: The Psychology of

Intuition and Its Role in Philosophical Inquiry, ed. Michael R. DePaul and William Ramsey, 113–27. Lanham, MD:

Rowman and Littlefield.

iv Tidman, Paul. 1996. “The Justification of A Priori Intuitions. “Philosophy and Phenomenological Research 56(1):


* Throughout this paper I will use the term perceptions to mean any insights that result from the physical stimulation

of sensing organs

v Nagel, Jennifer. “Intuitions and Experiments: A Defense of the Case Method in Epistemology.” Philosophy and

Phenomenological Research November 2012. 495-527.

vi Raftopoulos, Athanassios. Cognition and Perception: How Do Psychology and Neural Science Inform

Philosophy? Boston: MIT, n.d. 304-07. Print.

vii Russell, Bertrand. Our Knowledge of the External World as a Field for Scientific Method in Philosophy. Chicago:

Open Court, 1914. 52-55. Print.

viii Weinberg, J. M. (2007). “How to Challenge Intuitions Empirically Without Risking Skepticism.” Midwest

Studies in Philosophy, 31, 318–343.

ix Weinberg, J. S., Nichols, S. and Stich, S. (2001). Normativity and Epistemic Intuitions. Philosophical Topics,

29(1), 429–460.

11 Buckwalter, W. and Stich, S. (2011). Gender and the Philosophy Club. The Philosophers’ Magazine (52), 60–65.

x Koriat, Asher (2011). Subjective Confidence in Perceptual Judgments: A Test of the Self-Consistency Model.

Journal of Experimental Psychology: General, 140(1), 117–139.

xi Chabris, Christopher F., and Daniel J. Simons. The Invisible Gorilla: And Other Ways Our Intuitions Deceive Us.

New York: Crown, 2010. Print.

xii Kruger J, & Dunning D (1999). Unskilled and unaware of it: how difficulties in recognizing one’s own

incompetence lead to inflated self-assessments. Journal of personality and social psychology, 77 (6), 1121-34.

xiii Mercier and Sperber, D. (2009). “Intuitive and Reflective Inferences.” J. St. BT Evans and K. Frankish (Eds.), In

Two Minds: Dual Processes and Beyond. Oxford, UK: Oxford University Press, 149-176.

Leave a Reply

Your email address will not be published. Required fields are marked *