To speak of the history of philosophy is to speak of history; the history of philosophy is history, and so is subject to any generalizations which can be made about history.
Whichever descriptive or prescriptive rules apply to history in general apply also to the history of philosophy.
There is even the risk of an infinite regression: if we apply the philosophy of history to the history of philosophy, then we would have the philosophy of the history of philosophy - and we do in fact have this.
(And then: the history of the philosophy of history, the history of the philosophy of the history of philosophy, the philosophy of the history of the philosophy of history, etc.)
The term ‘meta-history’ can be used in several senses: it can refer to the archetypical patterns which either govern or describe historical events and processes; it can refer to the manner in which historical narratives are created; it can refer to the study of the philosophy of history (cf. the Oxford English Dictionary).
There is also the question of defining ‘history’ - the word can refer to the actual events and processes which have occurred in the past, or it can refer to our writings and narratives which attempt to record those events and processes, as Frederick Copleston writes:
Taken by itself, in abstraction from context, the term ‘history of philosophy’ is ambiguous. On the one hand it may refer to the actual historical development of philosophical thought, in the Greco-Roman world, for example, or in India. On the other hand it can be used, and, I suppose, generally is used, to refer to written accounts of such developments. Thus one might say of a book, ‘this is an account of the history of Greek philosophy.’ This possible ambiguity might be avoided by reserving the term ‘history of philosophy’ for the actual historical development of philosophical thought and by employing some such phrase as ‘historiography of philosophy’ to refer to written accounts of this development. But this cumbersome procedure is unnecessary. These sense in which ‘history of philosophy’ is being used is generally clear enough from context.
Copleston’s view includes some amount of Platonism or realism - he calls it ‘a common-sense point of view.’ This leads us to Kantian considerations about history.
Kantianly, there’s a Ding-an-sich (a ‘thing in itself’) behind or under our perception of the thing: a noumenon behind the phenomenon. Analogously, we might ask about an event in itself apart from the event as we perceive it.
On the other hand, an opposing view might argue that there is no history apart from the perceptions we have and the narratives we create: that history is the telling of history, and nothing more. Philosophers use the word ‘idealism’ to refer to such a view, and the Anglo-Irish philosopher George Berkeley (1685-1753) expressed such views.
If Copleston draws our attention the questions about realism or idealism in history in general, and in the history of philosophy in particular, then John Yolton draws our attention to the linguistic considerations:
Taking its root in the theory of types, the theory of meta-disciplines insists that any given discipline, such as sociology, psychology, language, etc., must be distinguished from the talk about that discipline. The psychologist in his laboratory, performing experiments, deriving conclusions, devising theories, is engaged in the accepted activities of his profession. When that same psychologist begins to reflect critically upon his methodologies, upon the meanings of the familiar terms, upon the general significance of his science, he has taken on the activities traditionally associated with the philosopher, he has, that is, become a meta-psychologist. Philosophies of science are, in this way, meta-scientific activities, since the men who construct them do not perform the scientific experiments but instead examine the presuppositions, the methodologies, the concepts of the sciences. Since the two-fold distinction of discipline and meta-discipline arose out of the theory of types, and since this theory has had its most wide-spread application within semantic systems, much of the activity of meta-disciplines has taken the form of a linguistic analysis of the speech and writing of any given discipline.
There is, perhaps, a general pattern in which questions which were viewed as metaphysical questions in the 18th and 19th centuries came to be viewed as linguistic questions in the 20th century.
If the 20th century was the heyday of the philosophy of language, then it should be no surprise that meta-historical thought was treated as a question about the language used to talk about history. Meta-history was seen as the examination of the language which historians used to talk about evidence and data - the language which was used to construct narratives.
Yolton continues:
Thus, Hare has recently given us an analysis of The Language of Morals, Woodger presents Biology and Language, and Ryle offers an account of mental conduct words. Such analyses of the language of disciplines cannot avoid making assertions or drawing conclusions about the activity of the discipline itself: that is precisely the function of a critical examination of science, ethics, or history, i.e., to formulate conclusions not only of a descriptive but also of a prescriptive nature. Ryle urges us, for example, to dispense with the fallacy of the ghost in the machine in thinking about the nature of mind and mental activity, while the historical relativist insists that claims to objectivity in historiography are fallacious. If taken seriously, the analysis of the meta-discipline reflects back upon the discipline itself, recommendations made upon the meta-level are absorbed 'into the original activity. The meta-discipline lays bare latent presuppositions, exposes them to criticism, challenges their accepted meanings, indicates contractions required for precision.
We might pose a question about whether, in the history of philosophy, it is correct to speak of Bossuet ‘reacting’ or ‘responding’ to Hobbes. One the one hand, to thus speak creates a more intelligible and continuous narrative. On the other hand, it might be more accurate to simply say that Bossuet wrote at a point in time after Hobbes had published his major texts. To attribute ‘reacting’ or ‘responding’ to Bossuet might exceed the evidence at hand.
Likewise, to ask whether George Berkeley is properly grouped among the British empiricists, as Louis Loeb does, is simultaneously to challenge a construct created by some historians of philosophy and to create a new competing construct. Both the asking of the question and the answering of it are worth investigating, inasmuch as the examination of schools of philosophy is the examination of a later historiographical construct, a construct which is then retrojected onto the earlier evidence, rather than a direct examination of the data itself.
In any case, while it is first necessary for us to naively ask a historian of philosophy for a narrative about “what happened,” it is secondly also necessary for us to ask the historian of philosophy about how he constructs narratives and how he chooses from among competing narratives.
