Genealogically, some languages may be attributed to a single source referred to as their proto-language. Similarly, dialects of the same language are related in a number of aspects but may differ phonologically.
An interesting aspect is that some phonological variations within language are non-continuous in space (i.e, distinct locations may share features while closer locations may show different features). This was been found for intonation contours, for example, and draws a picture unlike that found for segmental variation in most cases. The sources for this 'dispersion' may be various: some sound patterns may depend and/or interact with other aspects of grammar (e.g., , morphology, syntax, pragmatics); population drift within a country may also have a relevant impact; cultural issues can lead certain local communities to easily adopt the more general pattern or, by contrast, to persevere in the more local pattern. So, the array of factors is complex. We provided some illustrations in the Interactive Atlas of the Prosody of Portuguese (for intonation contours, phrasing, and rhythm features).
An interesting question, and as it happens I was involved in a related discussion just the other day, which stimulated some of the following thoughts.
I was thinking about the phenomenon of elisions of syllables and sounds in spoken English, and it struck me that the patterns of elisions tend to be distinctive in distinct dialects. But an additional point is that the semantic viability of elisions depends on the need of contrasting sounds to mark differences of meaning. In consequence, one might expect that patterns of elisions in a dialect are, in some degree, a function of drift in the vocabulary. Usage of some items increases over time and other decline. We expect that frequency of items of vocabulary will be related to the prevalence of specific activities and concerned in relation to which the vocabulary finds its use. As patterns of activities and employment shift for identifiable groups, then one also expects the relative frequency of items of vocabulary to shift and that in turn would effect the viability of elisions within the group.
Nothing I say here is intended to deny or question the various laws of sound shifts which have been proposed or documented. It is more a matter of other changes in language which may facilitate particular shifts in sound systems over time.
Modern means of communications have, in an important degree, decreased the role of local and regional dialects in many places, but they do still exist especially in the diminished form of local or regional sound patterns or "coloring."
There are many reasons for phonological variation in dialcts. One is that everyone learns language for themselves, starting in infancy. Toddlers don't mimic, indeed can't. Another is that, within a culture, when different ethnic or social classes are formed, each one distinguishes members by change in phonological and lexical differences. The way one speaks reveals his or her feelings of identity. In other words, social stratification results in direct differences. Also, if a group is admired, others will change their speech to be like them. My book, Language the Social Mirror, 4th ed., explains this phenomenon in depth. I can't upload it as it's still in press & is copyrighted, but you can probably buy it second-hand. Also, Amazon and Barnes & Noble sell it new.
Change and variation are so prevalent that I wonder if we might be better off asking the question from a different angle: What are the causes for non-variation in a speech community?
When asked this way, one see how hard it is to maintain status quo. Variation would really be the norm.
Indeed so, Lian-Hee Wee! Variation IS the norm - it's the starting point, if you like. Some linguists have a bad habit of thinking in terms of some standard form as basic and variation as some form of deviation from that norm. That's definitely the wrong way of looking at it. Standardization has developed out of the variation.
Hi David, when you refer to "phonological" variation, instead of the more known phonetic variation, do you mean 1) the same sound that produces different meanings in different dialects or 2) an advanced stage of phonetic change in a given lexical item, with a fixed meaning, common to both dialects, where an allophone gradually drifts toward something that (in other contexts!!) is a different phoneme (that is, recategorization or reanalysis)?
Clearly the second process (whose achieved resulting status, observed synchronically, is probably referrable to as phonological variation between the whole systems, in the sense of which pairs are phonological and which ones are not across dialects) may however lead to the first when the change of one form is able to hit another existing form causing polysemy and amibguity and if the newborn form prevails in use over its homophone and kicks it out of the vocabulary. So this process can actually be a motivation for phonological variation in the first sense I mentioned.
An interesting aspect is that some phonological variations within language are non-continuous in space (i.e, distinct locations may share features while closer locations may show different features). This was been found for intonation contours, for example, and draws a picture unlike that found for segmental variation in most cases. The sources for this 'dispersion' may be various: some sound patterns may depend and/or interact with other aspects of grammar (e.g., , morphology, syntax, pragmatics); population drift within a country may also have a relevant impact; cultural issues can lead certain local communities to easily adopt the more general pattern or, by contrast, to persevere in the more local pattern. So, the array of factors is complex. We provided some illustrations in the Interactive Atlas of the Prosody of Portuguese (for intonation contours, phrasing, and rhythm features).
There are cases where a phonological unit of a language corresponds to two phonological units in one of its varieties. The units appeared as a consequence of different processes in the history of the varieties. This may lead to the existence of different number of phonological units in the varieties.
The cases I'm familiar with show two phonological units that keep the same place of articulation, though they diverge in manner features. One variety has a more simetric inventory than the other, since the new phoneme has fulfilled a gap in the combination of manner and place features.
Phonetic variation exists in all language varieties. It is frequently related to the ease of the articulatory gestures, to coarticulation effects, to frequency of the lexicon used, etc., apart from all the factors reffered in the other interesting comments above.
There are no single causes for the emergence of new contrasts, I believe.
According to me, Phonological variations can occur due to the difference in lexicon, manner of articulation, prosodical variation, even variation in stress and duration also cause phonological variation in dialects and language varieties.