Language is fluid. Like the river Heraclitus stepped into, it is constantly changing and we cannot speak twice in the same language. However, in order to understand linguistic changes, we impose artificial parameters on language. For example, linguists recognize at least three distinct phases (thus far) in the history of the English language: Early English (EE), Middle English (MDE) and Modern English (ME). As anyone who has tried to read Beowulf will know, EE is extremely difficult for one fluent only in ME to understand. The time span for these major divisions of English use (beyond which it is difficult to recognize it as an immediate relation) is about a millenium.
It is difficult to pinpoint the exact point when one version of a language changes into an other. Changes are always occurring within a language and one must identify which changes are critical to distinguish an epoch change or, alternatively, how many changes need occur before an epoch change takes place.
Moreover, these changes appear at various levels and include phonemic changes, morphological changes, changes in the order and complexity of syntactic arrangement and ultimately, changes that can bring about the emergence of grammatical functions from proto-linguistic forms. At minimum, for a language version to be said to exist, some conditions must hold for a while, and maybe just long enough, for the language system to be shared in a population.
The rate of language change is difficult to establish, but we assume that smaller changes probably occur often and do not profoundly disrupt the shareability of language. In fact, we will suggest later on that changes promote the shareability of linguistic exchanges. Overall drastic restructuring in a language probably happens less frequently. Whether some of these changes happen incrementally or abruptly, the shareability of language is not hindered. Indeed, it is this shareability that defines the specific version of a language as a language and not another.
If a verse of Beowulf [42] is quoted in a conversation, it probably would not generate any specific effects, except perhaps for those who are familiar with the poem. In fact, it is very difficult to read without a translation. Nonetheless we can still recognize it, especially when heard, as a relative (if distant), of a ME. There is a clear boundary between our understanding, or lack thereof, of English spoken over a thousand years ago and now.
We inherit vocables generationally from our parents and our community; If we follow a diachronic account of these vocables, we notice that even though they may survive throughout several generations, their uses in a conversation may change substantially over time. Some changes in the language are quite dramatic. The Beowulf poem is a good example that much of the vocabulary is now incomprehensible to us. But we also observe that the survival rate of vocables is not shared equally amongst them. In Beowulf some vocables are more familiar to us than others. That is, the use, the morphology and sometimes the phonetic aspects of some vocables will remain somewhat unchanged for longer periods than others.
The Canterbury Tales are a good example of MDE. Though the text is readable, to a ME user, it is clearly not written in ME. When we look closely, most of the connective vocabulary has morphologically survived until now. But in examining some of the constructions it is clear that the instances of connective vocabulary in MDE are different from those of in present times.
And certes, if it nere to long to heere,
I wolde have toold yow fully the manere
How wonnen was the regne of femenye
By theseus and by his chivalrye;
And of the grete bataille for the nones
Bitwixen atthenes and amazones;
-Goeffrey Chaucer, The Canterbury Tales, The knight's Tale Part I- [12]
The use of for in And of the grete bataille for the nones is not the same as the use of for in say, this gift is for you. If we were to translate the first instance of for we would probably use because in its stead. The second use of for is a use of present times and is used in a context of benefactive. While the morphology of this bit of functionalized vocabulary has not changed, nor has it structural role in syntax, we can see that the benefactor for has its origins in the causal for. Sometimes subtle changes in language lead to more dramatic changes in the syntactic organization of sentences. Bernard H. Bichakjian has reported changes in language in the order of subjects, objects and verb use in the family of Indo-European languages. Bichakjian describes a reversal in syntactic constructions from head-first to head-last. He considers the heads of a sentence loosely as verbs and prepositions while their objects are the modifiers such as attributive adjectives. John's FOOT and the FOOT of the bed are examples in which the noun FOOT is the head. For each proposition, the head finds itself, respectively, last and first [7]. Bichakjian uses his observations to hypothesize that head-first language are more efficient than head-last ones, largely because they allow more complex constructions, particularly the use of embedded sentences.
We consider the occurrence of syntactic complexities, such as embedded sentences, a convergence of the erosion of specific effects generated from the use of lexical constructions. As the use of specific vocables becomes less rigid, we need to use more complex sentences to express ourselves. For all its seeming inconvenience, there are at least two reasons for the erosion of lexical uses: First, it can be, as Bichakjian suggests, exploited to increase linguistic efficiency. Second, language needs to adapt to environmental constraints. We will use the language of exploitation of incidental effects to describe the emergence of all kinds of vocables.
These are just a few examples that demonstrate a dynamic that starts in the everyday use of language where small unnoticeable changes occur. These unnoticeable changes are cumulative and eventually scale to very noticeable changes to the point of generating new versions of languages. Despite all these changes, thence or perhaps even because of them, the shareability of a language is never compromised.