In his youth, Berlin's intellectual development followed that of English-language philosophy, and he was at one point deeply involved in the advance of analytic philosophy; yet he drifted away from this, and his later writings and concerns are a world away from most Anglo-American philosophy of their time. Joshua Cherniss & Henry Hardy "Isaiah Berlin," Stanford Encyclopedia of Philosophy.
As a one-time avid reader of Isaiah Berlin I found nothing to disagree with the passage above written by extremely knowledgeable students of Berlin. I suspect, in fact, that if we looked at Berlin's later writings, he very much embraced this image of his relationship to analytic philosophy. While the passage is not false, I think it is seriously misleading. Berlin was a fierce critic of analytic philosophy.* In fact, he is completely up front about this in what is arguably his most famous bit of writing. I quote from the opening preamble of "Two Concepts of Liberty," (which I was re-reading in light of my exploration of Sophie De Grouchy's and Constant's analysis of the famous distinction):
If men never disagreed about the ends of life, if our ancestors had remained undisturbed in the Garden of Eden, the studies to which the Chichele Chair of Social and Political Theory is dedicated could scarcely have been conceived. For these studies spring from, and thrive on, discord... Where ends are agreed, the only questions left are those of means, and these are not political but technical, that is to say, capable of being settled by experts or machines, like arguments between engineers or doctors. That is why those who put their faith in some immense, world-transforming phenomenon, like the final triumph of reason or the proletarian revolution, must believe that all political and moral problems can thereby be turned into technological ones. That is the meaning of Engels' famous phrase (paraphrasing Saint-Simon) about 'replacing the government of persons by the administration of things', and the Marxist prophecies about the withering away of the State and the beginning of the true history of humanity. This outlook is called Utopian by those for whom speculation about this condition of perfect social harmony is the play of idle fancy. Nevertheless, a visitor from Mars to any British - or American -university today might perhaps be forgiven if he sustained the impression that its members lived in something very like this innocent and idyllic state, for all the serious attention that is paid to fundamental problems of politics by professional philosophers. ...
Our philosophers seem oddly unaware of these devastating effects of their activities. It may be that, intoxicated by their magnificent achievements in more abstract realms, the best among them look with disdain upon a field in which radical discoveries are less likely to be made, and talent for minute analysis is less likely to be rewarded. Yet, despite every effort to separate them, conducted by a blind scholastic pedantry, politics has remained indissolubly intertwined with every other form of philosophical enquiry. To neglect the field of political thought, because its unstable subjectmatter, with its blurred edges, is not to be caught by the fixed concepts, abstract models and fine instruments suitable to logic or to linguistic analysis - to demand a unity of method in philosophy, and reject whatever the method cannot successfully manage - is merely to allow oneself to remain at the mercy of primitive and uncriticised political beliefs. Berlin, I. (1958) “Two Concepts of Liberty.”
The main direct criticism of analytical philosophy is that its then-neglect of political philosophy is dangerous. For this neglect leaves dangerous political beliefs that exist unexamined and un-criticized. In offering this criticism, Berlin conceives of philosophy as an enterprise that ought to be judged on its social consequences and omissions. That is to say, philosophy is a moral vocation for Berlin.
Berlin's stance toward analytic philosophy is no different from analytic philosophers that object to the social consequences of, for example, Heidegger's philosophy or the criticisms of other projects that tend to be associated with authoritarianism and, say, forms of irrationally. We may understand Berlin's position as follows: analytic philosophers going back to Carnap pride themselves on our adoption of responsible speech (clear, rigorous, etc.) and have a tendency to cry 'charlatan' or 'fraud' when we encounter jargon-ridden continental philosophers, but we refuse to accept that our stance may generate omissions for which we can be held justifiable responsible. In particular, Berlin seems to believe that philosophers ought to guide society's ideas (if we don't such ideas "remain blind and undirected.") How this guiding is exactly supposed to happen he leaves somewhat unclear (he mentions "rational criticism") but he believes it is the duty of philosophers to "disarm" dangerous ideas before they become too dangerous.**
But this is not the only criticism (of analytic philosophy's stance & speech). Berlin also claims that analytic philosophy fools itself by understanding itself as a-political, despite the fact that "politics has remained indissolubly intertwined with every other form of philosophical enquiry." Attempts at separating philosophy and politics entirely are doomed to fail. Berlin is not entirely clear why this would be so, but the arguments for his position are not too hard to discern. Part of the position is, as we have seen, that philosophical practices generate systematic patterns of consequences and omissions for which philosophy can be held responsible. (Another version of this is Constant's endorsement of what I call a "Rhetorical Precautionary Principle" (recall; see also 'Socratic problem.')
A key aspect of Berlin's criticism of analytic philosophy is his rejection of the adoption of a technocratic stance. (This stance is more evident among then exiled members of the Vienna circle than in ordinary language philosophy, of course.) This criticism can be sub-divided in several sub-criticisms: (i) analytic philosophy mistakenly claims unity of methods (I use the plural because Berlin actually mentions more than one method); (ii) analytic philosophy is method driven, so that what the method can't handle cleanly (or with promise of success) it illegitimately rejects as a philosophical topic; (iii) technocracy presupposes or assumes agreement over ends, so that analytical philosophy takes as settled what ought to be an achievement. (Note that (iii) does not rely on Berlin's own tragic-tinted value-pluralism.)
From Berlin's perspective (iii) is an extremely serious charge because it helps generate "the impression that its members lived in something very like this innocent and idyllic state." That is to say, the more fundamental criticism of analytical philosophy is that by withdrawing into the purportedly responsible speech of technocrats, it fundamentally misunderstands the world it is in and, thus, endangers its own survival.
Berlin also has a technical criticism of analytic philosophy--one that is present in the passage that I quoted, but that also closes the essay (and, thus, frames it). It's the idea that concepts are unchanging. I think Berlin is willing to grant -- for the sake of argument, perhaps -- that some concepts are indeed eternal. But he thinks that such a view is dangerous, even "immature." The criticism is directed at (a) those social reformers that wish to impose abstract ideas onto particular societies without attention to local details (or in Constant's terms, the "spirit of the age") as well as (b) those philosophers that by embracing an understanding of concepts as unchanging withdraw from political philosophy--such philosophers are (and Berlin echoes Wittgenstein) not cured from a certain conception of metaphysics, even if they did so by claiming no interest in metaphysics (I quote the final paragraph of the essay):
It may be that the ideal of freedom to choose ends without claiming eternal validity for them, and the pluralism of values connected with this, is only the late fruit of our declining capitalist civilisation: an ideal which remote ages and primitive societies have not recognised, and one which posterity will regard with curiosity, even sympathy, but little comprehension. This may be so; but no skeptical conclusions seem to me to follow. Principles are not less sacred because their duration cannot be guaranteed. Indeed, the very desire for guarantees that our values are eternal and secure in some objective heaven is perhaps only a craving for the certainties of childhood or the absolute values of our primitive past. 'To realise the relative validity of one's convictions', said an admirable writer of our time, 'and yet stand for them unflinchingly is what distinguishes a civilised man from a barbarian.’ To demand more than this is perhaps a deep and incurable metaphysical need; but to allow it to determine one's practice is a symptom of an equally deep, and more dangerous, moral and political immaturity.-- “Two Concepts of Liberty.”
As we all know, analytical philosophy did recover an interest in political philosophy. So, much of Berlin's criticism feels a bit quaint, even if there are plenty of technocratic formal philosophers. But it remains true that even Rawls, who certainly recognized the need of conceptual engineering for society's benefit, embraced a technocratic conception of politics (recall):
By way of summing up, the essential point is that despite the individualistic features of justice as fairness, the two principles of justice are not contingent upon existing desires or present social conditions. Thus we are able to derive a conception of a just basic structure, and an ideal of the person compatible with it, that can serve as a standard for appraising institutions and for guiding the overall direction of social change. In order to find an Archimedean point it is not necessary to appeal to a priori or perfectionist principles. By assuming certain general desires, such as the desire for primary social goods, and by taking as a basis the agreements that would be made in a suitably defined initial situation, we can achieve requisite independence from existing circumstances. The original position is so characterized that unanimity is possible; the deliberations of any one person are typical of all. Moreover, the same will hold for the considered judgments of the citizens of a well-ordered society effectively regulated by the principles of justice. Everyone has a similar sense of justice and in this respect a well-ordered society is homogeneous. Political argument appeals to this moral consensus. (- See more at: https://digressionsnimpressions.typepad.com/digressionsimpressions/2014/02/the-technocratic-conception-of-politics-milton-friedman-stigler-and-rawls.html#sthash.HKs3394B.dpuf
In order to find an Archimedean point it is not necessary to appeal to a priori or perfectionist principles. By assuming certain general desires, such as the desire for primary social goods, and by taking as a basis the agreements that would be made in a suitably defined initial situation, we can achieve requisite independence from existing circumstances. The original position is so characterized that unanimity is possible; the deliberations of any one person are typical of all. Moreover, the same will hold for the considered judgments of the citizens of a well-ordered society effectively regulated by the principles of justice. Everyone has a similar sense of justice and in this respect a well-ordered society is homogeneous. Political argument appeals to this moral consensus. (Rawls, A Theory of Justice, 263)
If Berlin is right then (1) it is inevitable (because that's what technocracy requires)*** that Rawls appeals to such a moral consensus -- and in TJ Rawls essentially adopts a representative agent approach familiar from economics -- as well as (2) Rawls also has no right to this consensus (because it is constitutive of political philosophy that there is no such moral consensus [I leave the status of Political Liberalism aside here]). In fact, if Berlin is really right then (3) Rawls's path was a dangerous one, too.
*In fairness to Cherniss, his book, A Mind and Its Time, draws a far richer account of the opposition to analytic philosophy in Berlin's writings and many of the themes discussed in the post surface there, too.
** In context, he appeals to Heine, and so it is possible that Berlin believed such disarming of dangerous 'prophecies,' also required, in part, poetic means (see 'philosophical prophecy'). If that is so, then he has a further argument against the un-poetic technicratic stance of analytic philosophy.
***One complex feature of Berlin's position is that he seems to have recognized that if one accepts the demand for a formal semantics then value pluralism becomes much harder to defend.