REFERENCE BASE
LANGUAGE

O.T. FORD

Language is the most important form of communication, the use of symbols.

We observe the world, and attempt to notice patterns. We generalize from these patterns, and over time learn to connect the generalized patterns with meaning. A sign is a generalized pattern that we associate with a meaning. Most signs are direct interpretations of the way the world works: smoke is a sign for fire. Communication in general is an attempt by one person to elicit a certain interpretation by another. A symbol is a sign whose association between observed pattern and meaning is one of convention, where a particular pattern — a sound or a shape, for instance — could mean anything, but we choose for it to have a particular meaning.

Language also involves ordering rules, in which additional meaning comes from the arrangement of symbols. Taken together, the total of symbols and ordering rules understood by an individual is that individual’s idiolect. The intersection of the idiolects of a group of persons, the symbols and ordering rules they understand in common, is that group’s dialect. Dialects also include what are conventionally called “languages”.

For all hearing humans, speech is the primary form of language; it is learned first and becomes an organizing structure of our thought, as well as our communication. Speech involves symbols formed from sounds we make with our mouths. Written language, while important in literate societies, is subordinate to spoken language. We mostly use writing to represent language that exists primarily as speech.

Language evolves over time. Changes are introduced in small ways, by one or a few speakers, through exposure to new environments, development of new technologies, interaction with speakers of other dialects, individual preferences, and even sometimes as errors. Some of these changes become broadly adopted and eventually are typical for a dialect. An observable example of this is slang; slang emerges within a subgroup, creating new words or giving new meaning to old words; most slang dies out, but some catches on, becoming standard. This is normal for all language and happens continually. And if there were only one language group in the world, and if its speakers lived in constant communication, in a single village for instance, then there would only be one dialect in the world, albeit one that gradually changed over time.

However, language groups do not always stay intact, do not always keep in constant communication. When a group speaking a particular dialect splits into two groups, each of them will speak that same dialect in the beginning. But each dialect will change, as usual. If those two groups are not in communication, then each new change will not be communicated throughout the larger original group, but only through one of the new smaller groups. Eventually those changes will add up, and the two dialects will have changed so much that the groups cannot understand each other very well, or at all.

In traditional linguistics, a distinction is made between ‘a language’ and ‘a dialect’. In this usage, dialects of a single language are mutually intelligible, meaning that their speakers can understand each other, while speech varieties that are not mutually intelligible are separate languages. In reality, linguists have long recognized that mutual intelligibility is a matter of degree. No two persons understand each other perfectly. Many symbols are borrowed so widely that billions of people can understand them, even while understanding very little else in common. Linguists can only distinguish between “languages” and “dialects” by setting an arbitrary threshold for mutual intelligibility. Moreover, the distinction between “languages” and “dialects” is often made (occasionally even in linguistics) for non-linguistic reasons, particularly related to political power or national identity. Favored dialects are always “languages”; in a famous unattributed statement, “A language is a dialect with an army and navy.” (אַ שפּראַך איז אַ דיאַלעקט מיט אַן אַרמיי און פֿלאָט. « Ɔa šprɔak ɔīz ɔa dīɔalceqt mīt ɔan ɔarmīj ɔūn plɔot. »).

Dialects that evolved from the same source are said to be related; a group of related dialects is a language family. Typically the term is applied to the broadest known grouping, a grouping which is itself not known to be related to any other dialects. Smaller groupings within the family are then known as branches; each branch must have evolved from a single dialect which itself evolved from the original dialect of the entire family. While some of this evolution happened in the historical period (as Latin evolved into the Romance dialects, Spanish, Portuguese, French, Italian, and the like), most of it is reconstructed by linguists through comparison of dialects as they are spoken or have been recorded. The language families with the largest number of speakers:

      1. Indo-European
      2. Sino-Tibetan
      3. Niger-Congo
      4. Afro-Asiatic
      5. Austronesian

Indo-European dialects are spoken by nearly half the world. The family was originally identified through a comparison of Latin, Ancient Greek, and Sanskrit, and the study of Indo-European led to the basic methods of linguistic reconstruction. It must be noted that these families are based on the current state of the field; while Indo-European is widely accepted, some families are controversial, some internal relationships are uncertain, and there are proposals to group some of these families (including Indo-European) into even-larger families (for example, Nostratic), and at least some of those proposals are sure to be accepted in time.

The partition of language is highly geographic: it happens because of migration and loss of communication. A group speaking a single dialect breaks into two or more groups, at least one of which migrates to a different location, where regular communication with the other group(s) is no longer possible, given the technology of the time. It is conventionally viewed as resulting in discrete dialects and linguistic communities:

In this case, parts of the original dialect group have migrated to points A and Z, and there is no communication between A and N, or N and Z, because the distances are too great (and this is, of course, happening before the creation of modern communications and transportation technology).

But in most cases, what actually happens is the creation of a dialect continuum. A dialect continuum is a geographical phenomenon in which two or more related dialects, spoken in adjoining areas, blend into each other.

In this case, the original group spreads out over a larger area, but that area is continuous. Consider each lettered location a village; each village maintains communication with the villages immediately to the west and the east, engaging in trade, intermarriage, and the like. But the distances between A and N or N and Z remain too great for direct communication. Any changes in language — pronunciations, new words, new meanings — may spread from their points of origin to nearby locations, but not necessarily beyond. The constant communication between neighboring villages means that their dialects never lose mutual intelligibility, even if the dialects of distant points like A and Z are not mutually intelligible.

While we imagine the world divided up into discrete languages, the continuum is actually the norm. The map at right shows the South Slavic dialect continuum. The traditional convention has been to divide it into Slovenian, Serbo-Croatian, Macedonian, and Bulgarian; a more-recent convention further divides Serbo-Croatian, on political grounds, into Serbian, Croatian, and Bosnian. From a linguistic point of view, though, none of these distinctions is reliable, as there is no clear line between any of the vernaculars. (The division of Serbo-Croatian into dialects on linguistic grounds would produce a different result, with standard “Serbian”, “Croatian”, and “Bosnian” all part of the same Štokavski dialect, and not matching in any way the Štokavski subdialects of Ekavski, Ikavski, and Ijekavski.)

It should be noted that writing and sign language also evolve and also have family relationships, but not the same family relationships. The Brahmi family of scripts that originated in India is used for dialects of the Indo-European, Dravidian, Sino-Tibetan, Mon-Khmer, and Tai-Kadai families, while other Indo-European dialects are written in various scripts of the Semitic family of scripts — Latin, Cyrillic, Arabic, Greek. There are numerous dialects that are mutually intelligible but written in different scripts, such as Hindi and Urdu, or various dialects of Serbo-Croatian, and many dialects have been written in different scripts historically than is the case now. English and French are only distantly related through Indo-European, but American Sign Language is a recent descendant of Old French Sign Language.

The basic unit — the smallest symbol — of human language is often taken by its users to be the word, but many conventional words can be analyzed into smaller symbols, morphemes, which are the smallest possible meaningful units of language; the /s/ that marks a plural in English is an example. These morphemes are then assembled according to ordering rules to make longer symbols, which can be assembled according to further ordering rules, and so on.

Conventional words are classified as parts of speech — noun, verb, adjective, adverb, preposition, pronoun, conjunction. Fundamentally, though, there are only five basic functional classes through which ideas are expressed: nouns, which refer to things, verbs, which refer to actions (in a form analogous to ‘do’), modifiers, which describe things, actions, or other modifiers, dependent particles, which establish logical relationships among the other parts of speech, and independent particles, which convey complete thoughts without resort to the other four. Adjectives, adverbs, and articles are all modifiers, most pronouns are of course nouns, conjunctions and prepositions (and the conceptually-identical postpositions) are dependent particles, many interjections are independent particles. (So-called “possessive pronouns” are actually modifiers; many interjections are nouns, verbs, or modifiers.) These functional elements can be conventional words, but they can also be conventional phrases; there are ordering rules that allow complex nouns to be composed from other nouns, modifiers, and particles, or complex modifiers to be composed from nouns and verbs. A prepositional phrase, for example, is a modifier composed of a dependent particle and a noun.

Other important concepts:

— native dialect — the dialect an individual learns from birth, in its home, from its family. It is possible to learn more than one dialect from birth, and thus have more than one native dialect; but native dialects by definition cannot be changed, or acquired later in life. Native dialect is not racially or ethnically determined; it is not the original dialect of one’s group.

— vernacular dialect — the dialect used for everyday, casual communication, in the home, among family and friends.

— standard dialect — a dialect developed as a common form among many similar dialects, and for education and literature. Typically a standard dialect is simply the local dialect of an important place within a dialect continuum, such as the seat of government. It can also be a compromise among features present at different locations within the continuum.

— lingua franca — a dialect used to communicate between two persons or communities that do not share the same native dialect. Typically linguae francae are regionally or functionally based, so that, in a particular area or in a particular field of activity, a particular dialect will be used as the lingua franca. Linguae francae have historically been chosen based on two criteria: governmental power (the dialect of the seat of government or the ruling group) and trade (the dialect of an important trading location or population).

— official dialect — a dialect actually used in a state’s official business. This is not a matter of declaration, but observable fact.

— promoted dialect — a dialect which a government is attempting to spread, generally as an act of political nationalism. This is typically done through state education and through restrictions on other dialects.

— language isolate — a dialect with no known relatives. Basque, spoken on opposite sides of the Pyrenees in Europe, is a frequent example. It is highly unlikely that there are genuine isolates; more likely is that we have not yet established relationships that will eventually be known.

— pidgin — a new dialect that develops as a lingua franca in some instances where two groups without a common dialect encounter each other. Historically, the common pattern is that the dialect of the less powerful group will be the source of the basic structure of the pidgin, and the dialect of the more powerful will be the source of the bulk of the vocabulary. ‘Pidgin’ is itself a pidgin word, from English ‘business’ (as in ‘business English’).

— creole — a pidgin that has become a native dialect.

© O.T. FORD