John Dewey's conception of the role of criticism is quite straightforward and follows naturally from his conception of art:
The function of criticism is the reeducation of perception of works of art; it is an auxiliary in the process, a difficult process, of learning to see and hear. The conception that its business is to appraise, to judge in the legal and moral sense, arrests the perception of those who are influenced by the criticism that assumes this task. The moral office of criticism is performed indirectly. The individual who has an enlarged and quickened experience is one who should make for himself his own appraisal. . .The moral function of art itself is to remove prejudice, do away with the scales that keep the eye from seeing, tear away the veils due to wont and custom, perfect the power to perceive. The critic's office is to further this work, performed by the object of art. (Art as Experience, Ch. 13 (325)
One might prefer to think of the critic's task as simply the education of perception, although Dewey no doubt uses "reeducation" deliberately. So much of modern life inhibits the process of "learning to see and hear," making it all the more difficult than it is already given the influence of "wont and custom." Too often critics themselves work as impediments to clear perception, in particular those who intercede a "judicial" (Dewey's word for the approach to criticism that replaces explanation and analysis with a simplistic rendering of critical decision) and moralistic discourse between the work of art and those who need most to see and hear so they may finally judge for themselves. Both critic and audience need to be reeducated away from these habits.
For Dewey, a more useful form of "judgment" consists in distinguishing "particulars and parts with respect to their weight and function in formation of an integral experience." The critic must develop "a unifying point of view" with which to consider the work of art. However,
That the critic must discover some unifying strand or pattern running through all details does not signify that he must himself produce an integral whole. Sometimes critics of the better type substitute a work of art of their own for that they are professedly dealing with. The result may be art but it is not criticism. The unity the critic traces must be in the work of art as its characteristic. This statement does not signify that there is just one unifying idea or form in a work of art. There are many, in proportion to the richness of the object in question. What is meant is that the critic shall seize upon some strain or strand that is actually there, and bring it forth with such clearness that the reader has a new clue and guide in his own experience. (314)
While a critic of the "better type"--the type that is lauded for his or her own critical writing to the extent that it comes to take precedence over the writing under review--might be tempted to "judge" a work by comparing it to the work the critics thinks should have been produced but wasn't, the critic who sticks to the "object" (in literature, the text) actually in front of him/her is the one who is finally engaged in the act of criticism. The literary critic is obliged to honestly examine the characteristics the text exhibits, although he/she is not obliged to account for every characteristic that might be felt. The "unity" the critic posits is not a global unity that exhausts the work's formal or thematic possibilities but could be simply a "strain or strand" that does give the text coherence when shown to connect it's particulars in a satisfying way. As Dewey says, there are many such strands, "in proportion to the richness of the object in question," and one critic's analysis of "unity" can be supplemented by additional kinds of unity demonstrated by other critics.
In addition, the truly valuable critic avoids what Dewey thinks are the two "fallacies" of criticism. "Reductive" criticism occurs "when some constituent of the work of art is isolated and then the whole is reduced to terms of this single isolated element," or when the work is reduced to its historical, political, or economic circumstances. Dewey finds psychoanalytic and sociological criticism especially reductive. With the former, "If the factors spoken of are real and not speculative, they are relevant to biography, but they are wholly impertinent as to the character of the work itself." As to the latter:
Historical and cultural information may throw light on the causes of [the work's] production. But when all is said and done, each one is just what it is artistically, and its esthetic merits and demerits are within the work. Knowledge of social conditions of production is, when it is really knowledge, of genuine value. But it is no substitute for understanding of the object in its own qualities and relations. (316)
Thus academic criticism of the historicist and cultural studies varieties may result in something that could be called knowledge (although not always), but it is not knowledge of literature.
The second fallacy, "confusion of categories," can be related to the first when the critic fails to acknowledge this autonomy of the aesthetic. It happens when "critics as well as theorists are given to the attempt to translate the distinctively esthetic over into terms of some other kind of experience." The most common manifestation of this fallacy is the assumption
that the artist begins with material that has already a recognized status, moral, philosophic, historical, or whatever, and then renders it more palatable by emotional seasoning and imaginative dressing. The work of art is treated as if it were a reediting of values already current in other fields of experience. (318)
Thus the religious poet is declared to be the spokesman for a set of religious values, the philosophical poet for a particular philosophy, etc. But
medium and effect are the important matters. . .I imagine the majestic art of Paradise Lost will be more, not less admitted, and the poem be more widely read, when rejection of its themes of Protestant theology has passed into indifference and forgetfulness. . .The mise-en-scene of Milton's portrayal of the dramatic action of great forces need not be esthetically troublesome, any more than is that of the Iliad to the modern reader. There is a profound distinction between the vehicle of a work of art, the intellectual carrier through which an artist receives his subject-matter and transmits it to his immediate audience, and both the form and matter of his work. (318)
Protestant theology is Milton's "intellectual carrier." Paradise Lost is what it is, aesthetically. The literary critic who confuses these things, who allows the "carrier" to supersede "the intrinsic signifance of the medium" (319) is not a literary critic.