In reacting to my little book Language, Logic and Criterion (Amsterdam-Assen, 1971) Dr Barth sets forth some fundamental problems in philosophy. They are very important for every attempt to treat traditional philosophical problems with the help of modern logical methods. I would be very glad to make the following small contribution to the discussion. First of all I should like to make the main purpose of my book dear. According to me there is not one unique system in philosophy. Not only in the trivial sense that each philosopher has his own system, but also that there are many systems of logical empiricism for each philosopher individually. These systems vary according to the strength and strictness of their logical and empirical presuppositions. In a first preliminary approach we may state that the strictness and strength of each system are in an inverse proportion to each other, i.e. the stronger a system is (i.e. the more one can prove in that system), the less strict it is and vice versa. So one has e.g. systems based on intuitionistic logic, on classical logic, on dialectical logic, etc.; systems with and without modal logic; systems in which a reference to mystical experience is permitted and systems in which this is not so, etc. 1 So any answer to a philosophical question can only be given within a certain system. With the help of this method of admitting a variety of systems one can try to give as strict a foundation to each answer as possible. Whether one will commit oneself to a certain answer is another question. Of course, the stricter a system is, the better, but on the other hand: the more a system can explain the better it is. The strictest systems must leave many problems untouched. Whether metaphysics is possible depends on the system that is used. I hope to have shown in my book that these systems are respectable (i.e. that they have a sufficient degree of plausibility), but I do not pretend that I can prove that metaphysics is possible in every system. Where in the hierarchy of systems the 'respectable' systems (i.e. the systems with an acceptable degree of plausibility) ends and the doubtful systems begin is a matter of personal and/or common decision. A common decision is made by a Erkenntnis 9 (1975) 401--409.