protocol statement one of the statements that constitute the foundations of empirical knowledge. The term was introduced by proponents of foundationalism, who were convinced that in order to avoid the most radical skepticism, one must countenance beliefs that are justified but not as a result of an inference. If all justified beliefs are inferentially justified, then to be justified in believing one proposition P on the basis of another, E, one would have to be justified in believing both E and that E confirms P. But if all justification were inferential, then to be justified in believing E one would need to infer it from some other proposition one justifiably believes, and so on ad infinitum. The only way to avoid this regress is to find some statement knowable without inferring it from some other truth.
Philosophers who agree that empirical knowledge has foundations do not necessarily agree on what those foundations are. The British empiricists restrict the class of contingent protocol statements to propositions describing the contents of mind (sensations, beliefs, fears, desires, and the like). And even here a statement describing a mental state would be a protocol statement only for the person in that state. Other philosophers, however, would take protocol statements to include at least some assertions about the immediate physical environment. The plausibility of a given candidate for a protocol statement depends on how one analyzes non-inferential justification. Some philosophers rely on the idea of acquaintance. One is non-inferentially justified in believing something when one is directly acquainted with what makes it true. Other philosophers rely on the idea of a state that is in some sense self-presenting. Still others want to understand the notion in terms of the inconceivability of error.
The main difficulty in trying to defend a coherent conception of non-inferential justification is to find an account of protocol statements that gives them enough conceptual content to serve as the premises of arguments, while avoiding the charge that the application of concepts always brings with it the possibility of error and the necessity of inference.
See also EPISTEMOLOGY , FOUNDATIONAL- IS. R.A.F.