Ethan Christian Tan is a Second Lieutenant in the Singapore Armed Forces. In 2022 he will begin reading Philosophy, Politics & Economics at The Queen's College, University of Oxford. In 2020 Ethan was awarded the Grand Prize in the John Locke Institute Essay Competition
Is Intuition to Philosophy as Observation is to Science?
Science and philosophy share the fundamental goal of knowledge production. The former explains and predicts the material universe,[1] and the latter seeks to answer questions within the fields of logic, epistemology, metaphysics, and axiology.[2] This common objective will contextualise my discussion of observation and intuition as means to the end of producing scientific and philosophical knowledge respectively. Observation and intuition can both be thought of as independent sources of knowledge. When we observe, we receive information either via direct sensory perception or with the aid of instruments.[3] Intuition, on the other hand, is best defined by what it is not. An intuition does not result from inference, memory, testimony, reason, or observation;[4] of these sources, a knower still believes intuitions to be a priori truths [5] in the absence. The question of whether intuition is to philosophy as observation is to science can be broken down into three sub-questions:
What roles do observation and intuition play in scientific and philosophical knowledge production respectively?
How reliably do they fulfil these roles?
How exclusively do they occupy these roles?
These sub-questions correspond to the bases of comparison of role, reliability, and exclusivity respectively. The extent to which the set of answers for observation coincides with that of intuition is then the extent to which intuition is to philosophy as observation is to science. With this condition in mind, I propose the following answers. Firstly, observation and intuition both play evidential roles in supporting and refuting scientific and philosophical claims, and productive roles in producing them. They differ, however, in that evidential observation is used deductively when falsifying scientific knowledge claims, while evidential intuition lacks a deductive function. Secondly, both observation and intuition possess unreliabilities in precision and accuracy. While certain problems can be accounted for in both cases, observation and intuition still feature seemingly unresolvable issues with accuracy, which therefore threaten reliability. Lastly, the evidential and productive functions of both observation and intuition are non-exclusive, in that they can, in some cases, be substituted by other sources of knowledge. Thus, I contend that intuition is largely to philosophy as observation is to science, save for the fact that intuition, unlike observation, is not used deductively. What roles do observation and intuition play in scientific and philosophical knowledge production respectively?
In the production of scientific knowledge, the roles of observation are twofold. When applied to a pre-existing claim, observation is evidential, and can be used to support or falsify said claim. Observation can also precede and bring about the production of knowledge, in which case its function is productive. These roles are not mutually exclusive; observation can produce and later support the same claim. Observation is used as supporting evidence when corroborating a scientific hypothesis. That is to say, if a specific observation P (“This beaker of water boils at 100°C”) corroborates a general claim Q (“Water boils at 100°C”), we have increased confidence that Q is correct.[6] Moreover, as written by Hume, our confidence in Q tends to grow as more instances of P are observed.[7] In this way, through the corroboration of hypotheses, scientists use particular observations of the natural world to support general explanations of its behaviour. It is important, however, to differentiate between the deductive and inductive uses of evidential observation. To conclude with epistemic certainty that “all swans are white” from observing any number of white swans is to commit the fallacy of affirming the consequent, whereas the mere observation of one non-white swan would suffice to seductively falsify the claim.[8] Therefore, while observation can only serve inductively to support a hypothesis, scientists employ evidential observation deductively when falsifying them.[9] The productive role of observation is illustrated well by one of science’s most famous anecdotes. One summer, a young Isaac Newton observed an apple fall from a tree. He wondered "why... that apple [should] always descend perpendicularly to the ground", which would later inspire his work on gravity.[10] Science seeks to describe the natural world; the observation of a natural phenomenon can therefore produce scientific knowledge regarding said phenomenon. Does intuition share the evidential and productive functions of observation? The latter is a significantly less controversial question among philosophers—it seems that, at least sometimes, our intuitions do produce our beliefs. Many philosophers adhered to the justified true belief (JTB) account of knowledge, but adjusted their stances after learning of Gettier-type counterexamples, which produce the intuition that the justified true beliefs they describe do not count as knowledge.[11] If beliefs in the JTB account had caused intuitions regarding knowledge, then Gettier cases would have been dismissed as unintuitive. The thesis that intuition is used as evidence for and against philosophical beliefs is dubbed Centrality by Cappelen.[12] Opposing Centrality, Cappelen argues that the description ‘intuitive’, when applied to philosophical claims, can be removed or substituted without loss of meaning.[13] Ichikawa explains in support of this claim that we use intuitive propositions not because they are intuitive, but because they are true.[14] As Climenhaga notes, these objections are refuted by the fact that our intuitions tend to coincide with our beliefs, and that our confidence in a given belief does not decrease even with the awareness that it has been caused by our intuitions.[15] Centrality provides the best explanation for this phenomenon. Moreover, beyond treating their own intuitions as evidence, philosophers frequently 'explain away' others' intuitions when used to justify opposing claims or as counterexamples to their own; if intuitions were not seen as evidential, this would be unnecessary.[16] Like observation, intuition thus serves as evidence for and against philosophical claims. Examples of both cases are ubiquitous in philosophical dialogue;[17] other than the aforementioned Gettier problem, which is an instance of the latter, Searle uses the ‘Chinese room’ intuition pump to support his claim that machines cannot understand—they only simulate the ability to do so.[18] It appears, however, that evidential intuition is not employed deductively. Sound deductive claims rely on true premises, yet Gutting notes that philosophers tend to avoid this requirement with intuitions, neglecting to account for their truth and simply expecting their audience to share the same intuitions.[19] In many occasions, the deductive premise-conclusion structure is not even adhered to, philosophers instead supporting general claims with individual examples and counterexamples of intuitions.[20] Hence, this is where the evidential role of intuition diverges from that of observation: while observation is used deductively to falsify and inductively to support, the role of intuition is limited to that of support.
How reliably do they fulfil these roles?
Sources of unreliability can be split into two categories: imprecisions, referring to qualitative or quantitative inexactitudes or variations, and inaccuracies, meaning distance from the truth. The most apparent hindrance to the reliability of observation is the problem of imperfections in our senses and scientific instruments, classified as random and systematic error. 21] The former results from unpredictable changes in experimental conditions that usually affect precision, while the latter arises from consistent instrumental defects that typically affect accuracy.[22] account for these errors.[23] Nonetheless, scientists are able to analyse and mathematically account for these errors.[23] A more pressing inaccuracy arises in the theory-ladenness of observation: that our observations are always interpreted through the subjective 'lenses' of the beliefs, or theories, to which we subscribe,[24] limiting our ability to accurately observe a subject in its entirety. Kuhn outlines three types of observational theory-ladenness: Firstly, our beliefs can directly affect what we perceive. In an experiment by perceptual psychologists Bruner and Postman, subjects repeatedly mistook wrongly-coloured playing cards for their normal counterparts when viewed quickly.[25] This, according to Kuhn, depicts how our perceptions are affected by our conceptual resources.[26] Secondly, our beliefs also influence our semantic understanding of the descriptions of observations.[27] For example, a Newtonian physicist would interpret the word 'mass' to refer to a constant, while an Einsteinian would believe the value described by the same term to depend on the velocity of the object in question.[28] This prevents observers subscribing to different theories from truly understanding descriptions of each other's observations. Lastly, separate aspects of the same observation may vary in salience based on the theories of the observer. Kuhn proposes the thought experiment of Galileo and an Aristotelian physicist observing a pendulum. The Aristotelian, viewing the pendulum’s bob as falling under a string’s constraint, would have given attention to the bob’s weight, vertical displacement, and time taken to come to rest.[29] In contrast, Galileo would treat pendular motion as restricted circular motion, measuring the string's length, angular displacement, and time per swing.[30] Our intuitions, like observation, also appear to feature ‘imprecision’, or varying intuitions in response to the same stimulus. Experimental philosophers Petrinovich and O’Neill, when surveying participants, found that moral intuitions varied based on how a dilemma was phrased and the order in which different dilemmas were presented.[31] However, Zamzow and Nichols observe that people are less confident in their responsive intuitions to thought experiments when influenced, even unknowingly, by philosophically irrelevant factors.[32] Confidence, therefore, may be a mechanism to qualitatively determine the imprecision of our intuitions, reminiscent of the calculable experimental uncertainty of observations. In terms of accuracy, Kahneman suggests that like our observations, our intuitions are also influenced by our beliefs and theories that are 'trained' by our environment.[33] A professional chess player can be said to act on highly accurate intuitions during a game due to the regular environment of chess and repeated practice within it.[34] Contrarily, it is difficult to call philosophy a regular environment: philosophical intuitions span a huge range of topics, and may sometimes extend into non-philosophical domains that provide contextual knowledge.[35] Neither can philosophers reliably practice intuiting, as the accuracy of corrective feedback is hard to determine due to the general lack of consensus in the philosophical community (as opposed to chess, which has easily analysable games and binary win-lose outcomes).[36] Like observation, intuition thus also features an accuracy problem, in that it is difficult to define standards by which the truth of intuitions can be determined in the first place. How exclusively do they occupy these roles?
Although observation is a heavily utilised evidential source for most of the empirical sciences,[37] exceptions exist in abstract disciplines like theoretical physics. Ellis and Silk criticise string theorists and proponents of the many-worlds interpretation for relying on the "elegance" and explanatory power of their theories, rather than empirical observation, as supporting evidence for their veracity.[38] Since their theories describe unobservable phenomena, an observation also loses the ability to falsify their claims.[39] Regardless of the normative question of whether observation ought to be used as scientific evidence, that the scientific community has, in fact, appealed to alternative sources of evidence shows the non-exclusivity of observation’s evidential role. Describing non-observational causes of scientific knowledge production may be more difficult as many scientific claims fundamentally result from observation; even theoretical claims like string theory seek to explain observed phenomena.[40] Accounts exist, however, of other unconscious processes producing knowledge claims, such as August Kekulé's dream of an ouroboros inspiring him to describe the ring-shaped structure of benzene,[41] which was only confirmed directly by observation in the following century.[42] The emergent field of experimental philosophy illustrates well the non-exclusivity of intuition’s evidential role in philosophy. Beyond intuition, contemporary evidential sources have expanded to encompass reaction times,[43] neuroimaging,[44] and human behaviour,[45] using observation, in fact, to support philosophical claims. Experimental philosophy also features alternative sources of philosophical knowledge production. Schwitzgebel and Rust questioned if ethics professors behaved with higher or more consistent moral standards than their non-ethicist colleagues; only after collecting behavioural data could they formulate the claim that there is generally no variation between ethicists' and non-ethicists' moral behaviour.[46] As a whole, therefore, observation and intuition do not exclusively fulfil their evidential and productive functions.
Of the three bases of comparison—role, reliability, and exclusivity—intuition and observation are similar on all counts, with the exception that intuition, unlike observation, does not play a deductive role in philosophy. Thus, I conclude that less the aforementioned difference, intuition is to philosophy as observation is to science.
*********
For a full list of footnotes and a full bibliography, please see below: