Points to know about the Neogramarians
The Neogrammarians' main contention was that language change was systematic: for example, the fact that Latin pisces (fish), tenuis (thin), and centum (100) correspond to the OE words fisc, þyn, hund(red), means that, normally, any word beginning in p-, t- or c- (=k-) in Latin would have f-, þ- or h- if it appeared in Old English. This is because the Indo European voiceless stops p,t,k remained unchanged in Latin, but became voiceless fricatives in Germanic:
Their critics pointed to the large number of apparent exceptions to rules such as these, but the Neogrammarians replied that where the rules seemed not to work, this was always due to one of the following:
a) further rules, perhaps not yet discovered, needed to be taken into account (example: Verner's Law)
b) the workings of analogy needed to be taken into account (example: the word for "five" in many Indo-European languages has taken the same initial letter as "four": Old English feówer, fíf, Latin quattuor, quinque, but Indo-European *kwetwer,*penkwe.
c) the word in questions had been borrowed from another language, and so had not undergone the sound-changes concerned (Example: Latin caseus, OE ciese 'cheese')
Family tree model
The Neogrammarians' concept of language change provided for a Stammbaum ('tree') theory, which assumed that languages were related to one another and branched out from a common stem. For instance, Indo-European was the old parent, the main families such as Italic, Germanic, Celtic or Slavic were the children, and their offspring the grandchildren. From this concept we get the terms 'sister', 'daughter' and 'mother' languages. One of the earliest exponents of this concept was August Schleicher in (Die Deutsche Sprache, 1868).
Although we still use this concept today, it has to be taken with reservations. Few modern philologists accept the idea of a single tribe of people speaking a single common language (Indo-European) which later split into various families of languages as the tribe drifted apart. Indo-European was almost certainly a family of widely-ranging dialects from the beginning. Splits between languages do not occur sharply - languages slowly diverge as dialects, and later their differences may become great enough for them to be called separate languages. But dialects can also converge (grow more alike) and blend together or affect each other in complex ways.
Even after they have split from each other languages may continue to influence each other on various levels.
And quite unrelated neighbouring languages may also influence each other. Words are often 'borrowed' from other languages (loans), and grammatical forms and pronunciations can also move between languages. Thus the uvular r has spread out from French into German and Danish, and both French and German share the same use of the past perfect (have + past participle) to mean the simple past. Johannes Schmidt published Die Verwantschaftsverhältnisse der indogermanischen Sprachen in 1872, where he explained the existence of similar features in neighbouring languages by his Wellentheorie, or Wave Theory of language innovations which spread from single point out over surrounding dialects.
While the wave theory sees linguistic characteristics spreading by means of normal contact, large-scale linguistic changes often occur as a result of more violent upheavals.
When two languages are forced into close contact because of conquest or mass population movements, so that they are both spoken in the same territory, they may interact with each other in various ways.
If the new language is taken up by the older inhabitants, they may introduce charactaristics of their old speech into the new language. This is called the Substratum effect. It is often said that the phonology of Irish English is influenced by the Irish language. Irish, in this case, is the substratum, the underlying layer.
Another possibility is that the new language becomes the politically dominant language, the language of power and administration, but it is not adopted by most of the population, who keep on speaking their old language. This happened in England after the Norman occupation, when French became the politically superior language. One result of this was that large numbers of French words were taken into the English language, and the old English words were discarded. Sometimes this happened because the new words expressed new concepts, sometimes because the Engllish speakers just liked using them - they were "smarter" or "posher". In Icelandic this is called "að sletta". The technical term is relexification.
It can also happen that two different language groups occupy the same territory in relative peace, intermarrying and eventually becoming one. This seemed to have happened in the areas in the North and East of England as a result of the Norse invasions - the Norse and the English spoke very similar languages, and in time they fused to become one. This may have speeded up language change quite a lot - English was already losing its inflectional endings, and Norse had different inflectional endings - the confusion might have contributed to the loss.