In a couple of earlier posts in this series, I've said that a distinction between matters of fact and matters of opinion is treated in Bishop Whately's Elements of Rhetoric. He doesn't devote space to this distinction in his logic textbook (Elements of Logic), though he does have some things to say about matters of fact (pp. 218-34). His brief discussion there again treats matters of fact as pertaining to particular, observable phenomena. In fact, he speaks of matters of fact as constituting the 'data' (p. 222) on which reason operates. (Today, it looks like 'facts' in its old sense has largely been replaced by the term 'data'.)
Whately was a major presence at Oxford, where he and his books influenced many students, including John Henry Newman. Whately's rhetoric text was used for many years at Harvard, where Thoreau was required to master the book. According to one author, Elements of Rhetoric was one of the two 'most widely used rhetoric texts in American colleges between 1835 and 1865, going through at least fifty-one American printings'.(Robert Connors, Composition-Rhetoric: Backgrounds, Theory, and Pedagogy, 1997, p. 61) So, Whately's way of distinguishing between matters of fact and matters of opinion must have been widely known in the 1800s.
Other 19th-Century textbooks took up the fact-opinion distinction primarily (or exclusively) in connection with testimony as a source of evidence. For example, here's a passage from John Bascom's Philosophy of Rhetoric (1866):
An easy fallacy of testimony is confounding facts with conclusions drawn from them. For competent testimony to the first, actual observation is all that is requisite. Men, though by no means equal in their power to observe or report events, are so relatively. The value of an opinion, on the other hand, depends almost exclusively on him who gives it. False conclusions, therefore, disguised under the facts, and stated as a portion of them, may readily embarrass testimony and mislead the judgment. A man is said to be insolent, to be proud, to be angry, to be drunk. It is a question, not purely of facts, but of inference from words and actions. (Bascom, Philosophy of Rhetoric, p. 74, 1870, 1st published 1866)On p. 75, Bascom adds this example: 'A patient testifies to the effect of medicine as if his statement were the explicit delivery of facts, and not the most uncertain of all opinions.'
Henry Philip Tappan's brief discussion of the fact-opinion distinction appears on pp. 426-9 of his Elements of Logic (1844):
By opinion we mean a judgment of the mind, respecting a proposition as true or false. Opinion is to be distinguished from absolute knowledge, as implying that the proposition which is its object, is still debateable. Testimony cannot establish the truth of opinions or judgments. Their truth can be established only on some necessary principle of the Intelligence. Testimony, as evidence, relates merely to matters of fact. All, therefore, that a witness can testify to, in relation to opinions, is the fact that he or some other person entertains such and such opinions. But the truth or falsity of the opinions must be determined on other grounds, and wholly independently of testimony. A man may be of the highest integrity, and of sane mind, and may sacrifice reputation and possessions, and life itself, in maintaining his opinions, without affording any evidence of their truth. His testimony only goes to establish the fact that he believes the proposition in question, and that he believes it ardently and firmly. (Tappan, Elements of Logic, 1844, p. 426)Here's Tappan on facts:
By fact, we mean phenomena [sic], — something which we know by observation merely. Facts are of two kinds: 1. Facts of the Senses, or external observation. 2. Facts of the Consciousness, or internal observation. By truth, we mean that which is arrived at by the pure Reason. We always assume observation as conditional to the exercise of Reason. But while observation supplies facts, Reason supplies the principles under which the facts are to be reduced. Now, whatever the Reason supplies, whether in intuition or in deduction, we call truth. From this comparison of truth and fact, it must still more clearly appear that testimony cannot prove truths or doctrines. Testimony is only an attestation of what has been observed. Truths or doctrines can be proved by reasoning alone. (Tappan, Elements of Logic, 1844, p. 427)Adams Sherman Hill offered a more critical treatment in his Principles of Rhetoric (1878). Just before quoting a legal text's definition of 'matters of fact', Hill writes:
[T]oo much stress is often laid on the distinction between matters of fact and matters of opinion, — since opinion enters into almost all statements with regard to matters of fact; since the instant an individual fact is doubted upon reasonable grounds its existence becomes matter of opinion; and since doubtful matters are those with which argumentative composition chiefly deals. (Hill, Principles of Rhetoric, 1878, p. 201)Hill then adds:
The real distinction is between matters into which fact most largely enters, and those into which opinion most largely enters; for, though the honesty of a witness is hardly ever the only thing to be considered in determining the value of his testimony, yet in some classes of questions his intellectual character tells for much more than in others. (Hill, Principles of Rhetoric, 1878, p. 202)So, in the 1800s, university students in English-speaking universities were likely to have had some acquaintance with a fact-opinion distinction. Still, as I noted in the penultimate paragraph of an earlier post, and as the above quotations confirm (and as at least one other blogger has noticed), labeling something as an opinion was not tantamount to exempting it from rational support or critique or from the sphere of true or false claims. Indeed, as Bascom suggests (above), to call something an opinion implied that the holder of the opinion should, if anything, be held to a higher standard of intelligence and skill in reasoning (compared to a witness testifying about a matter of fact) if her/his opinion was to be taken seriously. This connotation is preserved in one of today's uses of 'opinion'; specifically, it's preserved in our speaking of an expert's opinion (e.g., a medical opinion).
Here's a rough hypothesis: today's fact-opinion distinction (esp. as taught in secondary schools) blurs elements of the 19th-Century rhetorical distinction with the fact-value distinction that would later be emphasized in social sciences. Two other factors are at play. First, there's the legitimacy of a fact-opinion distinction in the context of journalistic ethics (where both terms are given sharper, more technical interpretations). Second, there's the frequent use of 'opinion' as a translation of the ancient Greek term 'doxa'.
I'm no master of ancient Greek, but from what I understand, 'doxa' differed significantly in meaning from the English word 'opinion'.* Crucially, doxa was often set in opposition to 'epistemê' (knowledge, or science), and doxa sometimes took on a very negative value, especially in works by Parmenides, in some of Plato's writing, and in Hippocrates' 'Law'. That last ancient source includes a claim that has been translated as follows:
There are in fact two things, science and opinion; the former begets knowledge, the latter ignorance. (Hippocrates, Vol. 2, trans. W. H. S. Jones [New York: G.P. Putnam's Sons, 1923])Those lines from Hippocrates (in assorted translations) are repeated all over the place, on syllabi, motivational sites, critiques of psychiatry, Royal Society publications, an opinion piece by a Congresswoman in an academic journal, young-adult fiction, an IEA conference site, etc. The sentiment behind uses of the quoted lines is admirable. Still, we shouldn't forget that while 'opinion' is there used to render 'doxa', the translated term was an exotic creature with a far stranger life and range of associations than 'opinion'. We shouldn't try to shoe-horn 'opinion' into the role of the old Greek term by pretending that opinion is somehow antithetical to knowledge or (even more of a stretch) to modern science.
(I say 'even more of a stretch' because 'doxa' was sometimes applied by the ancients to beliefs derived from sense experience, the source of empirical data.)
*I recommend the entry on 'doxa' by Barbara Cassin and Charles Baladier in Dictionary of Untranslatables: a Philosophical Lexicon, ed. Barbara Cassin, trans. Rendall, Hubert, Mehlman, Stein, and Syrotinski, trans. ed. Emily Apter, Jacques Lezra, and Michael Wood (Princeton University Press, 2014 [first published in French in 2004]), pp.228-30.