This is a remarkable and far-reaching book written with impeccable scholarship and considerable acumen. The timeline of its study begins in the late 1890s with Russell and Moore’s declaration of their “New Philosophy”. It continues through developments in their thought and the contributions of Stout, Whitehead and Wittgenstein. Ramsey’s writings in the 1920s, at once synoptic and iconoclastic, conclude the study. The connecting thread concerns how the early analytic philosophers’ evolving conception of the particular-universal distinction both influenced and was influenced by their evolving conception of analytic philosophy. It was a process of mutually beneficial illumination in which Cambridge was the crucible of the analytic enterprise. The agenda facing these philosophers was to select between the options provided by two orthogonal distinctions. Ontological pluralism and ontological monism differ about the number of entities there are, whereas categorial monism, categorial dualism and categorial pluralism differ about the number of categories.
One of the many refreshing elements of the book is how it upends much conventional wisdom about the origins of analytic philosophy. For instance, it selects Kant rather than Frege as the progenitor. By raising the question of what the categories are, the forms of representation essential for thought, Kant called into question the concepts of substance and attribute and the relation between them. His Metaphysical Deduction was an ill-fated rescue mission to save these and other “pure concepts of understanding”. What remains from Kant, and was taken up by Moore, was the conviction that the particular-universal distinction was indissoluble: either both sides obtain or neither. Switching to the formal mode, a predicate is what is predicated of a subject whereas a subject is the subject of predication. Again, Moore is well known for rejecting Hegel’s ontological monism but MacBride convincingly shows that Moore’s “The Nature of Judgement” (1899) endorses categorial monism: that in taking concepts to comprise the only category, he rejected the particular-universal distinction. Russell bolstered this endorsement by drawing on Bradley and arguing that the idea of a substance is misconceived since the parent idea of something determinately referred to and described by a subject-predicate judgment is itself untenable. MacBride very effectively mines Russell’s unduly neglected The Philosophy of Leibniz (1900) for these and other early statements in the “New Philosophy” of how and why the subject-predicate framework is to be abandoned. Perhaps Russell’s most striking argument runs: The surface form of language is misleading, for we can as well say “Humanity belongs to Socrates” as “Socrates is human” and in each case we express the same judgment. “Humanity” may not be the grammatical subject of the second quoted sentence, but humanity belongs just as much to the subject matter of that sentence as Socrates does. Russell’s philosophical development involved considerable turbulence: in The Principles of Mathematics (1903) he rejected this argument against categorial dualism, only for the argument subsequently to be refurbished by Ramsey and deployed against him. MacBride sees the “New Philosophy” as having a revolutionary phase followed by a reactionary one, as Russell and Moore lapsed from advocacy of categorial monism to apostasy. What ensued was an extended episode of whack-a-mole in which periodic resurgences in Cambridge of the particular-universal distinction, often in a non-standard form, were subjected to strenuous criticism by Stout, Whitehead and Ramsey.
I noted that MacBride displaces Frege with Kant as the ur-source of analytic philosophy. MacBride also makes clear how much progress Russell and Moore made independently of Frege. This is especially evident in their appreciation of the structural significance of relations. Entities stand in different relations only if those entities are distinct and so, Russell and Moore inferred, numerical distinctness is not to be accounted for in terms of relational difference. Moore went further: if there are only universals, from this \(f\) being over here and that \(f\) being over there, it does not follow that there are two \(f\)s. It follows only that \(f\) recurs—that \(f\) is over here and over there. To secure bearers for such universals, and to safeguard our ordinary judgment that there are two \(f\)s, Moore invoked the category of particulars, understood as instances of universals. Categorial dualism was thereby reinstated. Ordinary objects were identified with clusters of property-instances, allowing the possibility of distinct property-instances that are themselves exactly alike.
By 1911, however, Moore found the particular-universal distinction to be unclear. One consideration is that the supposed category of universals is gerrymandered, containing monadic properties, relational properties and relations. This realization opened up the possibility of categorial pluralism: that there is no a priori limit to the number of categories. Beginning in 1905 and working independently, Whitehead also drew this conclusion. By rejecting a bifurcation of nature between primary and secondary qualities or of what is observable and what is an instrumental posit, he permitted a diverse ontology, a host of entities that apparently share no common features and so belong to a multiplicity of categories.
During the 1910s and 1920s Stout embraced ontological pluralism and categorial monism on a posteriori grounds: what we perceive is a wealth of property-instances (where property-instances form a single primitive kind of entity) that are unified in concrete or distributive ways. These concrete ways yield ordinary particulars and the distributive ways yield determinate or determinable qualities. Universals are eschewed. Stout and Moore subsequently locked horns over the nature of property-instances. In MacBride’s opinion, Moore had the better of the argument. First, ordinary speakers do not draw upon a grasp of the theory of classes to understand predicative sentences. A fortiori, they do not understand “the glass is fragile” in terms of the glass having a property-instance that is a member of the class of fragile things. Second, to say that the glass and the spider web are fragile is to predicate the same thing of them. Distinct particulars, however, cannot have the same property-instance, although they can have the same universal.
From 1903 Russell’s understanding of the particular-universal distinction evolved as his thoughts changed about both propositions and relations. Russell abandoned realism about propositions because of the problem of accounting for false propositions. He came to treat talk of propositions as a mere way of speaking, thanks to his multiple relation theory of judgment and his conjecture that judgment relations have significant higher-order structure. MacBride argues that Russell retained that theory up until 1919 (through his logical atomism phase), reviving his earlier view that each non-symmetric relation has a “direction”. To deflect Wittgenstein’s famous 1913 criticism of the theory, Russell reverted to a quasi-Fregean view that concepts (i.e. universals) were exclusively predicative and incapable of serving as logical subjects. Notably, Russell recognized that belief ascriptions have different logical forms according to what is believed. There is then no constraint on the number of categories that might be involved in the content of a belief and the way to categorial pluralism is open.
In 1906–1907 Russell devised a proto picture theory of language that inspired the Tractatus; Russell’s doubts about his own theory were addressed by Wittgenstein’s more developed efforts. Moreover, the theory that propositions are pictures of reality and that complex propositions are truth-functions of elementary propositions dispenses with the particular-universal apparatus. This liberation movement reached its apogee with Ramsey’s incisive contributions. For Ramsey, there is no a priori reason why the language required for expressing and characterising atomic facts will be anything like predicate calculus. For example, there might be not just two but three or four or \(n\) different modes of basic grammatical combination; the overlapping capacities of individual expressions to combine with one another may confound any binary distinction.
MacBride’s book investigates a microcosm and a macrocosm. The microcosm is the particular-universal distinction. The macrocosm is a debate between monism, dualism and pluralism: a debate about how many ontological categories there are. Could the macrocosmic debate have been engaged with even if a different microcosmic debate had arisen? Maybe, instead of the particular-universal dualism, a different venerable dualism could have been dominant in early analytic philosophy: the dualism of mind and body. Is it then a historical accident that the debate about the particular-universal distinction assumed the significance that it did? If Russell had written The Analysis of Mind (1921) some thirty years earlier, would the debate between Cartesian dualism, idealism and neutral monism have secured foundational status? The particular-universal distinction was central because it is related to questions about the forms of judgement, questions emphasised by Kant, and also to questions about the forms of sentences, questions elevated by the linguistic turn. Nevertheless, had Kant not usurped Descartes, the nature of mind could have remained pivotal. Perhaps the fact that philosophy of mind took centre stage in philosophy in the 1970s and 1980s was redressing an imbalance, albeit resulting in an imbalance of its own.
MacBride has written an absolutely first rate study of early analytic philosophy. The clarity of his writing, the carefulness of his elucidations, the brilliance of his metaphysical discussions, as well as his sympathetic approach to the writing of the Cambridge philosophers, mark out this important and profound work.