Discipline of art, being a field of humanities, inherits from it two fundamental foundations: the historical character of the main directions of research and the expression of the results of all research without exception in the form of text. The first leads to the total dependence of any attempts at theorizing on existing generalizations based on the specifics of empirical material, the second to the absence of full-fledged theoretical knowledge, which, as is known, can only be expressed in the form of formal-logical constructions. Ultimately, both lead to a very conditional status of discipline of art as a scientific discipline. This situation is taken for granted, ontologized, and reproduced in each new generation of art researchers. In examining the reasons for this, three aspects are identified, a change in attitude towards which can qualitatively transform the disciplinary status of discipline of art. Firstly, it is the original character of cognitive activity, which is transferred from the initially empirical to the abstract level. Secondly, it is the subject of cognition, which may not be specific works interpreted as art, but either the research activity itself in relation to them in the entire breadth of the amplitude of its possible development, or what these works are regardless of their concretization. And, finally, thirdly, these are methodological foundations, which, in connection with the redefined nature of cognitive activity and the subject of cognition, must be replaced by fundamentally different ones: the activity approach to cognition, the method of deobjectification, the genetic-constructive approach, the logic of predicates used as a method of formalization. As a result, a situation is formed in which two models of the existence of discipline of art as a full-fledged scientific discipline are possible. In the first of them, the art criticism research activity itself in relation to art turns out to be cognizable, and thus modernized discipline of art acquires the character of a metadiscipline in relation to traditional historical discipline of art. The second model is based on strict formal-logical constructions, fixing not what was or is, but the entire possible amplitude of what is possible, and, consequently, when it appears, it is precisely what is discovered in the specifics of empirical material in the conditions of traditional historical art criticism research. It is important that both models imply the conclusion of the consideration of the visual (audio, audiovisual, conceptual) from under any mediating discourses: in the first case, translating the study into the measurement of interrelated ontological schemes of deobjectified texts, in the second case, into the measurement of formal object-predicate constructions. These conclusions, affirming the fundamental principles of the potential existence of discipline of art as a full-fledged scientific discipline – defining the vectors of its possible qualitative transformation, have a specific applied significance for it.