Serious attempts have also been made to establish scales for the measurement of relative intelligibility among immediately related idioms, [1] but here also it becomes difficult to establish consistently firm ground. It is generally accepted that degrees of inter-intelligibility are influenced by a wide variety of extraneous factors, including the acuity of hearing, relative intelligence and previous linguistic experience of the person whose comprehension is being judged, as well as reciprocal feelings among speakers of immediately related languages, and the subject-matter actually under discussion.

It must be once again stressed that there can be no absolute definition of any of the three terms outer language, language or dialect, since their application depends not only on degrees of relative inter-intelligibility but also the relative complexity of any given linguistic environment. Thus although both [52] Deutsch (German) and [79] Han-yu (Chinese), in all their respective varieties, are presented in terms of these three layers of immediate relationship, it is clear that Chinese embraces an even wider range of internal variety than does German.

Another significant feature of the linguasphere, clearly apparent from the Register, is the way in which small isolated languages more often than not contain deep divisions within themselves, as though the voices of each internal community need to be able to define themselves in terms of their closeness to voices of at least one other related, but distinct, form of speech.

In Europe, for example, such internal division is found within such isolated and relatively “endangered” ethno-linguistic entities as:

 [40] Euskara or Basque (with nine inner languages and over thirty dialects, extending across the Pyrenees from Vizcaya to the Pyrénées-Atlantiques);
 [41] Saame or ‘Lappish’ (with three outer languages, nine inner languages and fourteen further dialects spoken across northern Scandinavia and into Russia);
 [50] Gaeilge+Gàidhlig or Gaelic (with a sequence of five inner languages and up to thirty dialects spoken from the south coast of Ireland to the northern Hebridean islands of Scotland);
 [51] Rumantsch+Grischun with Nones+Cadorino or Ladin (with two outer languages and eight inner languages spoken in the Alpine valleys of Switzerland and Italy);
 [52] Frysk+Frasch or Frisian (with four outer languages, ten inner languages and double that number of dialects - spoken on or near the North Sea coast and islands, from the Netherlands to Schleswig-Holstein); and
 [53] Serbska+Serb__ina or Sorbian (with two inner languages and a dozen dialects, spoken in a small area south of Berlin).

For an exemplification of the way in which the Linguasphere classification has been applied to the detailed configuration of languages in other areas of the world, see any of the individual zones of the Register, and particularly the discussion in many of the headings to individual zones. A synopsis of the zones within each sector is presented in Volume Two (pp. 16-35).

Finally, it should be mentioned that a potential by-product of the Register’s listing of inner languages within outer languages relates to the planning and application of machine-translation. It is obvious that the problems confronting the design of a bilingual translation program are considerably less when the program is translingual (i.e. covering languages classified within the same net), and especially when the two languages are immediate enough in relationship to be treated within the same outer language. In other words, help in the development of published literature in a previously unwritten or little written language can be expected to benefit from the machine-translated version of literature already existing elsewhere in the same outer language, depending of course on the availability of a programmer with the necessary linguistic knowledge. Research in Mexico has already led to the creation of computer programs which can move a text largely by substitution between two very closely or immediately related languages, in contrast to the general reformulation required by translations between unrelated or distantly related languages. [2]