Richards was that Cambridge professor of criticism who turned literary criticsm upside down in the 1930's. He inspired the New Criticism and won the admiration of poets such as T.S. Eliot. Trained originally in psychology, Richards penetrated into a new level of hard-headed thinking to literary criticsm, pushing through the effusive waffling of critics past. Richards' work dealt mainly with poetry and in short, his burning question is what makes a poem great.
Richards dismisses all visual imagery from legtitimate poetic criticism. The conjuring of mental images is an uncontrollable process. Ask 10 different people what visual images are evoked by a line and you will get 10 different answers. Such images are more often than not, biographical with respect to the reader. Furthermore, the ability to visually associate with words varies tremendously from person to person. As such, it is a useless criteria to judge a poem with visual imagery in a group context.
Indeed, Richards argues that for criticism to be legitimate, it must concern itself with things that can be experienced in the same way by different people. Talk of things that vary from person to person is useless. This point is so central that Richards literally defines a poem as a group of words that evokes a particular experience that does not vary greatly when read by different sensitive readers. Furthermore, the experience depends crucially on the sequential arrangement of words.
The emphasis on experience may seem to be excessively abstract. However, Richards chooses the high road of meaning as the starting point of poetry because people would otherwise concentrate on irrelevant concrete details such as rhythm and rhyme. Concrete technial features like rhythm are fine in a poem but it is hardly what makes a poem interesting. As interesting thought experiments, Richards takes lines from famous poems and substitutes them with prosaic and nonsensical lines that bears the same rhythm. As you can imagine, the substitutes do not sound particuarly poetic.
It is the meaning of the words that determine the success of rhyming and rhythm. Richards proselytises against the schools of literary criticism that hold the form as the paragon of poetry. Without the idea behind them, the form itself become a meaningless cage, all the more dazzling because they are empty of essence. There is nothing particular enobling about the sonnet form, or the iambic pentamer. The haiku is no more mysterious than the rhyming couplet. Rather, it is what past poets have tried to say within these forms that have made them great.
Still, this is not to say that poetic devices are unimportant. Otherwise, there would be no difference between prose and poetry. In his definition of a poem, Richards specifies that in a poem, an invariant experience is evoked through the use of, amongst other things, the sequencial ordering of words. In prose, the sequence of words is relative unimportant as long as the meaning is conveyed. In poetry, on the other hand, the relation of words further back in the poem exerts an almost magical influence on later words to create new patterns of meaning. This rich insight owes much to Richards' training as a psychologist. Richards' argues that readers have an innate psychological tendency to look for patterns in a sequence of words - whether it be patterns in rhyming, scansion or rhythm. When one is reading prose, this tendency is normally repressed whereas in poetry, this tendency is exploited. When a line is read, one has a expectation that something similar will occur. When something similar does follow, aural associations are made and simultaneously, meaning associations are also made.
Richards argues that poets exploit this psychological tendency to look for patterns in order to reinforce or create new layers of meaning in a poem. Poetic techniques are used to control the experience of the reader in the reading of the poem. Long slow syllables will reinforce statements of doom. Sharp, clattering syllables bolsters descriptions of actions. Other techniques create new meanings through suggestion and association. As readers automatically look for rhyming patterns, when rhymes do occur at the end of a couplet, the words that rhyme are given new association. Such words, when appropriate, create a new web of meaning over with the prosaic meaning of the line.
Given the subordination of technique to meaning, Richards argues that the worth of a poem lies first and foremost in its meaning. One must first ascertain the meaning before judgement can be meted out. Once the expereience has been grasped, then judgement can be made on the worth of the experience itself - the profundity of the thought, the originality of the thought, or the concreteness of its evocation.