A closely related concept is the analytic language, which in the extreme case does not use any inflections to indicate grammatical relationships (but which may still form compound words or may change the meanings of individual words with derivational morphemes, either of which processes gives more than one morpheme per word).
Isolating languages are in contrast to synthetic languages, where words often consist of multiple morphemes. That linguistic classification is subdivided into the classifications fusional, agglutinative, and polysynthetic, which are based on how the morphemes are combined.
Although historically languages were divided into three basic types (isolating, flectional, agglutinative), these traditional morphological types can be categorized by two distinct parameters:
- morpheme-per-word ratio (how many morphemes there are per word)
- degree of fusion between morphemes (how separable words' inflectional morphemes are according to units of meaning represented)
A language can be said to be more isolating than another if its correspondence between word and number of morphemes approaches 1:1 more than the other one.
To illustrate the relationship between words and morphemes, the English word "rice" is a single word consisting of one morpheme only (also "rice"). This word has a 1:1 morpheme-word ratio.
The English word "handshakes", "hand-shake-s", on the other hand, is a single word consisting of three morphemes (namely, hand, shake, -s). This word has a 3:1 morpheme-word ratio. On average, words in English have a morpheme-to-word ratio substantially greater than one.
It is perfectly possible for a language to have one inflectional morpheme yet more than one unit of meaning. For example, the Russian word vídyat (видят) 'they see' has a morpheme-per-word ratio of 2:1, having two morphemes: the root vid-/вид-, and the inflectional morpheme -yat/-ят which inflects for four units of meaning (3rd person subject, plural subject, present/future tense, imperfective aspect). Effectively, four units of meaning in one inseparable morpheme: -yat/-ят.
Languages that are relatively more isolating have a morpheme-per-word ratio that approaches 1:1. A purely isolating language would lack any visible morphology, since no word would have an internal compositional structure in terms of word pieces (i.e. morphemes) – thus it would lack bound morphemes like affixes.
The morpheme-per-word ratio is a scalar category ranging from low morpheme-per-word ratios (near 1:1) on the isolating hypothetical pole of the scale, to a high morpheme-per-word ratio. The greater a language's morpheme-per-word ratio is, the less isolating and the more synthetic it is.
See also 
Further reading 
- Sapir, Edward. www.bartleby.com/186/6.html Types of linguistic structure. In Language: An introduction to the study of speech Chapter 6. (1921).