In this collection of sixteen of his early papers, mainly translated from Polish, German and French, Leszek Kolakowski discusses Luther, Spinoza, Pierre Bayle, Uriel de Costa, Pierre Gassendi, Hegel, Marx, Avenarius, Nietzsche, Heidegger and Althusser. In the foreword Kolakowski says there is no ‘common theme’ in the book (p. vii) but most of the essays are concerned inter alia with the possibility of theology. All are written in an accessible style and are full of human interest. His book is a pleasure to read.

Kolakowski is rather harsh on Spinoza. The Spinoza of the title ‘miserably failed’(p. 14) in his philosophy because he did not avoid inconsistencies over, for example, freedom and determinism, the existence and non‐existence of God, science and mysticism, toleration and stability in politics. Kolakowski diagnoses the bifurcation between a German ‘metaphysical’ interpretation of Spinoza and a French ‘political radical’ interpretation in these inconsistencies. However, many great philosophers must be grossly mistaken in their solutions to philosophical problems because they disagree with one another. (For example, at most one of Hobbes and Berkeley can be right about what there is.) If Spinoza's claims are not only false but contradictory he perhaps deserves special chastisement but the systems of, say, Plato, Kant or Hegel are difficult to interpret as entirely internally consistent.

Kolakowski is right to be sharply realist about solutions to philosophical problems: If I am free then it is within my power to not do what I do. If my actions are causally determined then I cannot but do what I do. If there is a God then it is false that there is no God and vice versa. If everything knowable is scientifically knowable then there is no knowledge accessible only through mysticism and vice versa, and so on.

In ‘The Philosophical Role of the Reformation: Martin Luther and the Origins of Subjectivity’(pp. 143–160) Kolakowski calls subjectivity ‘the embryo of modern philosophy’ and says ‘philosophy is constantly striving to return to a primary, unmediated human subjectivity’(pp. 159–160). I contest this. Although Husserl's doctrine of the transcendental ego admittedly falls under this description, it is an exception. For Kant, subjectivity is formally constituted by the transcendental unity of apperception. For Hegel, subjectivity is socially constituted at a profound level by the struggle of master and slave. In scientific and pseudo‐scientific philosophy there is no subject, or only a reduction of the subject to a complex physical object. In poststructuralism the subject is deconstructed. If Kolakowski's embryo grew to be Cartesian it was aborted soon thereafter. It does not make much sense to speak of the ‘origins’ of subjectivity unless these are divine. One's own existence qua one's own is a metaphysical mystery that cannot be explained away, or even explained, philosophically.

Kolakowski rightly criticises Louis Althusser for a lack of analytical rigour in his For Marx(1969); grossly and tendentiously assimilating ‘says’ and ‘proves’ for example, and for huge historical blunders, such as ascribing a quasi‐Aristotelian or Scholastic theory of abstraction to empiricists. It would be interesting to hear Kolakowski's judgement on Althusser's paper ‘Ideology and Ideological State Apparatuses’ in Lenin and Philosophy and Other Essays(1971). At the end of ‘Karl Marx and the Classical Definition of Truth’(pp. 173–195) Kolakowski says Antonio Gramsci's interpretation of Marx's epistemology is ‘roughly in line’ with his own and arguably Althusser's allocation of causal efficacy to ideology in historical transitions is partly anticipated by Gramsci.

In ‘Heresy’(pp. 263–288) Kolakowski claims ‘A historian cannot accept the definition of heresy accepted in the Roman (or any other) Church, otherwise he would be assuming the viewpoint of a particular body, and the teaching of this body would be decisive in identifying the historical facts’(p. 266). Although there is such a thing as not assuming the viewpoint of a particular body there is no such thing as writing history without deploying some set of assumptions. History is more explanatory if methodologically self‐conscious, so if the historian's assumptions are Catholic they should be made explicit as such. It is the responsibility of the historian to write the truth, to report what happened in the past as it happened. Suppose the Roman Catholic definition of ‘heresy’ is correct. It follows that those doctrines correctly identified as heretical by the Church really were heretical. If the historian should write the truth, he should write that truth. It is historically impossible for the historian to deploy a retrospective epoche which guarantees agnosticism about beliefs held in the past, because the historian is himself historically situated. History is a relationship between the present and the past, or one time and another.

It is not true that by eschewing a Catholic commitment the historian occupies some ‘neutral’ vantage point. There is no such thing as not being committed.