§ 76. =The Snares of Language.=—You were warned on p. 36 that language may be misleading, and that the phrases which you naturally use oftentimes imply a view of the world, or an attitude towards experience, which is foreign to science. Nowhere, perhaps, is this discrepancy greater than in the phrases which refer to the self. Language, as we know, is older than science, and expresses the results of common-sense interpretation rather than of factual observation. _The self of language is, accordingly; not the psychological self, but the counterpart of the mannikin-mind_ (p. 7); and just as we must be on guard, and remember our psychological definition, whenever in a psychological context we say or think the word ‘mind,’ so must we be on guard against the common-sense notion of ‘self’ that has insinuated itself into a thousand turns of familiar speech. An observer, describing a particular experience, may say, quite naturally, ‘I find no trace of self-reference!’—and there is no harm done, if we realise that the _I_ of his remark is the traditional self-concept of language, and the _self_ the psychological experience of self; but there may be very great harm, if likeness of words leads us to confound the personal with the impersonal, common sense with science. Only by an unreadable pedantry can we avoid the I-phrases and the other personal sentences; but we must always bear in mind that _language, the very form and structure of it, embodies a theory, an explanation or interpretation of the self_; and that, if we reject this theory, we have to couch our criticism in terms of the theory we reject.