This is not an introduction to the subject of linguistic form, but a brief commentary on current views and terminologies.
If it is possible to discover any aim common to all linguistic schools, this aim is the reduction, by terminological devices, of the fundamental asymmetry of linguistic systems. If there are phonemes, allophones, and phonemic components, then there must also be morphemes, allomorphs, and morphemic components. If there is a form and a substance of the expression, then there must also be a form and a substance of the content. ·u every phoneme can be split up into a set of relevant phonic features, then every morpheme can be split up into a set of relevant semantic features. If the syntagm has a binary structure, so also does the syllable.
The relation between these four views (of which perhaps no two are held by the same scholar) is unmistakable.
There is hence perhaps room for a work that seeks to stress the fundamental asymmetry of linguistic systems, rather than reduce it. However, parallels between different levels have been noted whenever they seemed legitimate or at least suggestive.
Questions of historical linguistics have been left for treatment elsewhere, and references have been reduced to a minimum since a select critical bibliography is being prepared for publication in technically more advantageous circumstances. What remains is of course a mere sketch. But space has been found to touch on several aspects of linguistic form which pass unnoticed in more extensive treatments of the subject. On the other hand, the problem of segmentation, which claims almost a monopoly of attention in such treatments, has for this reason been left in the background.