We might say that a belief is something that, at least in principle, can be shown to be either true or false. It does not make sense to say that an opinion is true or false, because people differ greatly in their preferences and tastes. ... Here, an effort will be made to use beliefs to represent our ideas about what is objectively true, and opinions to represent such subjective matters as preferences and tastes. (Nickerson, Reflections on Reasoning [Hillsdale, NJ: Lawrence Erlbaum Associates, Inc., 1986], p. 21)Granted, Nickerson's claims have an air of stipulation. Still, why limit opinions to taste? Previous posts show that 'opinion' hasn't historically been defined in terms of taste. Indeed, before the 20th century, authors generally kept tastes out of the extension of 'opinion'. For example, here's a bit from Hume's 'Of the Standard of Taste' (1757):
Every voice is united in applauding elegance, propriety, simplicity, spirit in writing; and in blaming fustian, affectation, coldness and a false brilliancy: But when critics come to particulars, this seeming unanimity vanishes; and it is found, that they had affixed a very different meaning to their expressions. In all matters of opinion and science, the case it opposite: The difference among men is there oftener found to lie in generals than in particulars; and to be less in reality than in appearance. An explanation of the terms commonly ends the controversy; and the disputants are surprised to find, that they had been quarreling, while at bottom they agreed in their judgment. (italics added)In the first part of this passage, Hume is talking about taste (about which 'critics' dispute). He then turns to 'opinion and science', which are not matters of taste. His use of 'matters of opinion' as being different from matters of taste seems sensible in view of the meaning of 'opinion'. In his time, 'opinion' had come to mean something like a belief for which the justification was less than certain. The opinion would concern some matter of fact (on our more modern use of 'matter of fact'); e.g., it might concern whether all ideas originate in the senses. So, opinions were true or false, even if there was no consensus about their truth value.
Hume's use of 'matters of opinion' fits with the definitions of 'opinion' in today's general dictionaries (addressed in an earlier post). For example, the most recent edition of Webster's has it that an opinion is 'a belief, judgment, or way of thinking about something'. Hume's use is also in synch with the definition of 'opinion' in the American Psychological Association's Dictionary of Psychology, in which an opinion is said to be 'an attitude, belief, or judgment' (2007, p. 648).
Hume wasn't alone in keeping tastes out of the domain of matters of opinion. Consider these sentences by Jeremy Bentham:
There is no incident imaginable, be it ever so trivial, and so remote from mischief, from which this principle [of sympathy and antipathy] may not extract a ground of punishment. Any difference in taste: any difference in opinion: upon one subject as well as upon another. No disagreement so trifling which perseverance and altercation will not render serious. (Bentham, An Introduction to the Principles of Morals and Legislation, ch. 2, sec. 16 )Bentham is making a general claim about all disagreements. To cover such a wide range, he conjoins differences in taste with differences in opinion, implying that the former are not already included among the latter.
In recent fact-opinion distinctions, it has been customary to categorize expressions of taste as matters of opinion or even to equate opinions with tastes (as Nickerson does). I maintain that before the 20th century, matters of opinion were not equated with, or even taken to include, matters of taste. Matters of opinion concerned beliefs that were true or false and that were subject to logical principles. By contrast, rightly or wrongly, many authors from the 18th century to our own time have subscribed to the adage de gustibus non est disputandum ('there is no disputing about tastes') and have accordingly regarded matters of taste as lying outside the domain of logic and truth claims.
It's difficult to support my contention by finding old (pre-20th-century) examples of authors explicitly distinguishing between matters of opinion and matters of taste. So, I'll have to rest content with the less direct support that I've taken from the above quotations of Hume and Bentham, supplemented by two quotations from less famous sources. First, in his study of the New York Public Library's history, Tom Glynn says that near the end of the 19th century, the library's administrators took the view that 'there is a great distinction between matters of opinion and matters of taste' and that the library's holdings should include materials that represented both sides in debates over conflicting opinions (Glynn, Reading Publics: New York City's Public Libraries, 1754-1911 [New York: Fordham University Press, 2015], p. ???). Next, there is a line in a memorial to a Unitarian preacher, James Freeman, where the author says that Freeman 'carried [his] freedom of mind into matters of taste as well as matters of opinion' (James Freeman Clarke, Annals of the American Unitarian Pulpit , p. 169). The author is compelled to mention matters of taste because they are not included among matters of opinion.
At some point in the 20th century, authors began to include tastes among matters of opinion and even, in some cases, to use 'matters of taste' interchangeably with 'matters of opinion'. For example, in a 1958 article, David Peace writes that '"Matters of taste" are matters of opinion' (Peace, 'Planning and the Promotion of Good Taste' The Journal of the Royal Society for the Promotion of Health 78 : pp. 339-40).
Similar views are evident in empirical research about when children are able to distinguish between statements of fact and of opinion. In one study, the authors say, '3-year-olds acknowledge and understand disagreements between individuals about matters of taste and opinion more readily than disagreements about matters of fact' (Larisa Heiphetz et al. 'What do Different Beliefs Tell us? An Examination of Factual, Opinion-Based, and Religious Beliefs' Cognitive Development 30. April - June 2014 : 15–29). Here, statements of taste and opinion are lumped together as being mere subjective expressions that must be distinguished from statements concerning objective matters of fact (claims about what's true independently of us). In the 'Appendix' to Heiphetz et al., all of the 'statements of opinion-based belief' that were used in their research are expressions of taste (e.g., 'Oranges are the tastiest fruit', 'Monopoly is the most fun game'). Something similar happens in this report about an article by G. P. Goodwin and J. M. Darley ('The psychology of meta-ethics: Exploring objectivism' Cognition 106 [March 2008], pp. 1339-1366). (One sometimes gets an impression of similar uses of 'matters of opinion' in the literature on faultless disagreement, though contributors to that literature are generally more careful about the phrase.)
Even if we set aside its deviance from the pre-20th-century use of 'matters of opinion', this recent equation of matters of opinion with matters of taste is puzzling. After all, one today hears occasionally about opinions as including primitive prejudices that stand in the way of science. That seems to account for the popularity of this Hippocratic line: 'There are in fact two things, science and opinion; the former begets knowledge, the latter ignorance.' (For more about this quotation, see the end of this post.) But if opinions were mere tastes, then how could they obstruct science? After all, claims about what is the most fun or the tastiest don't pose any apparent threat to scientific data or theories. Clearly, if opinions are to include anti-scientific prejudices, they can't be mere expressions of personal taste; they must involve (wrong-headed) truth claims about something other than tastes.
Somehow, the phrase 'matters of opinion' slipped its moorings and drifted away from the standard meaning of 'opinion'. This slippage has prompted some authors to consider the possibility that many opinions (even those that do not count as knowledge) do not concern matters of opinion (Reuben Stern and Sam Wilkinson, '“That’s Just Like, Uh, Your Opinion, Man”: Jesus, the Dude, and Ordinary Language Philosophy', in The Big Lebowski and Philosophy: Keeping Your Mind Limber with Abiding Wisdom, ed. Peter S. Fosl [Wiley, 2012], p. 244).
As an antidote to this frankly screwed-up change in the use of 'matters of opinion', I recommend Michael Scriven's 1971 contribution to the Diablo Valley Education Project in Orinda, California. Scriven's paper is called 'Values and the Valuing Process' (pdf). Scriven there writes, 'Matters of opinion are issues where the possibility of the opinion being right or wrong exists; matters of taste are not' (p. 3).
There's also a 2009 psychology paper in which the authors try to make sense of some recent data by distinguishing between matters of opinion and questions of taste. They say, 'we define matters of taste as essentially arbitrary preferences that therefore have no intrinsic power or prestige associated with them, whereas opinions (like beliefs) refer to views that can at least in principle be evaluated at some level for their correctness' (Russell Spears, Naomi Ellemers, and Bertjan Doosje, 'Strength in Numbers or Less Is More? A Matter of Opinion and a Question of Taste' Personality and Social Psychology Bulletin 35 : pp. 1099-1111).