Languages In the World (50 page)

Paleontologist Simon Conway Morris (2011) has said, “Once there were bacteria, now there is New York” as a way to sum up the obvious increase in complexity since the origin of life on this planet. Researchers in a range of contemporary sciences – from physics to biology to economics to linguistics – have circled around the topic of complexity and tried to understand the seemingly general phenomenon that highly unstable, dynamics systems – such as the cosmos, living beings, cultures, and languages – become more complex over time and that the complexity is irreversible. When dealing with stable and undynamic systems, the concept of complexity is not problematic. If one is counting and comparing the number of parts in various objects, then one can confidently say that a computer is more complex than an abacus. Transferring this quantitative understanding to the realm of biology, one measure of complexity might reasonably be the degree of hierarchy found in various organisms as they develop over time: prokaryotic cells without nuclei came first; they were followed by eukaryotic cells with nuclei; it took more time to develop complex multicellular organisms and then even more time for certain organisms to organize themselves into even more complex colonies. The historical trajectory of this biological nestedness is a clear case of an increase in complexity over time.

Turning to matters of language, we have some ideological clearing to do before we can move forward. In the case of the topic of complexity, a twofold clearing is in order. The first clearing came around the turn of the twentieth century when anthropologically oriented linguists, in particular those studying Native American languages, worked to break the old association between linguistic complexity and technological complexity. Previous generations of European and American intellectuals had subscribed to a tripartite cultural and linguistic global hierarchy with so-called Barbarians at the bottom, Savages in the middle, and Civilized People at the top. The civilized people were, of course, those who had the most technologically advanced materials, namely all those things listed in the Final Note in Chapter 2 that the English and Spanish brought with them to the New World. Franz Boas, Edward Sapir, and Leonard Bloomfield were among the early twentieth-century linguistic anthropologists to show that Native American languages were every bit as evolved as European languages in terms of expressive power and structural complexity. Other linguists came to the same understandings of indigenous languages in Australia, Africa, and other technologically undeveloped places. In fact, it was noted that some so-called primitive peoples spoke languages with amazingly intricate structures.

A truism took hold in linguistics that all languages were equal in terms of complexity. Call this
equilibrium linguistics
, and it entails a view that the same degree of complexity will be found in all languages in terms of their states at any given time and in their states over time. In comparing two languages at their present states, Hawaiian and
English, we observe that Hawaiian has a phonemic inventory of 18, while English has double that number at 36. The supposed phonemic simplicity of Hawaiian is, however, offset by its 11-way personal pronoun system, which is more complex than the mere five of Standard English. Thus, in their current states, it can be said that Hawaiian and English are equally complex, although their complexities are found in different places.

When comparing states of the same language at two different times, a similar equilibrium of complexity is supposedly at work. Old English had an amount of inflectional morphology that Modern English speakers find completely baffling, exemplified by lines 4 and 5 of
Beowulf
:

 

The indirect objects of the main verb
ofteah
‘took away' are in the dative plural case. The object of a transitive verb would normally be in the accusative case. However, in this case, it is in the genitive, which makes
ofteah
a verb of the genitive of depriving. What? The Modern English speaker wonders: Why all the unnecessary morphological complexity to say something as simple as
Often Scyld Schefing took away many enemies' mead benches
?
¹⁰
On the other hand, Old English had only two tenses, present and past, while Modern English has a greatly expanded verbal system. Old English speakers would be perplexed how to understand (and by the need for!) the verbal sequences in an utterance such as
She will have already completed her homework before I even begin to start mine
. The inflectional complexity of Old English is gone, but the tense/aspect complexity of the Modern English verb system is in full swing. It would seem that in the passage of time, English has had neither a net gain nor a net loss of complexity.

The second ideological clearing has just now begun at the start of the twenty-first century, and it was creolist John McWhorter who shot the first salvo across the bow of equilibrium linguistics with his article entitled “The world's simplest grammars are creole grammars” (2001). We said above that unstable, dynamic systems seem to increase in complexity over time and that the complexity is irreversible. Creoles present an interesting case in that they are language loop reboots. We said at the end of Chapter 8 that whatever the relative exposure to the standard lexifying language that a generation of children creolizing a pidgin had, there is necessarily a strong break with the standard lexifying language loop as a whole, in that languages are coral reef accumulations of habits of mind and cultural practices. In our terms, we would say that McWhorter is right to say that it takes time for a language to loop in and around itself, and the ways that language does so are what we call
linguistic structure
. So, a language just starting down its path will likely have fewer loops, a looser weave so to speak, than a language that has been on the road for millennia. Creoles offer cases of what happens when the arrow of time is interrupted, and the complexity of the lexifying language is reversed in the born-again creole.

McWhorter's study of 19 English-, French-, Dutch-, and Portuguese-based creoles show that they all share a similar grammatical profile, and none of them have features found widely distributed in languages with long histories such as the ones listed in the above subsection and which McWhorter identifies as:

… ergativity, grammaticalized evidential marking, inalienable possessive marking, switch-reference marking, inverse marking, obviative marking, ‘dummy' verbs, syntactic asymmetries between matrix and subordinate clauses, grammaticalized subjunctive marking, verb-second, clitic movement, any pragmatically neutral word order but SVO, noun class or grammatical gender marking (analytic or affixal), or lexically contrastive or morphosyntactic tone beyond a few isolated cases. (2001:163)

Since none of these more elaborated features are strictly necessary for communication, they can be seen as what happens over time when the language keeps looping around itself. McWhorter calls these loops “baroque accretions” and the “weight of ‘ornament' that encrusts older languages.” The oldest of the creoles in use today are not very old. They are the Portuguese-based creoles of Cape Verde, spoken on an island in the middle of the Atlantic Ocean, and Guinea-Bissau, spoken on the west coast of Africa, both of which date back only to the late fifteenth century. The French and Dutch Caribbean creoles trace back to the late seventeenth and eighteenth centuries, while the English creoles of the Pacific go back to the eighteenth century. Hawaiian Creole English emerged only toward the end of the nineteenth century. In the terms of this book, young creoles have not yet confronted as many conditions to catch up to as have older languages.

McWhorter defines complexity in quantitative terms. A phonemic inventory is more complex the more members it has. A tonal system is more complex the more tones it has. A syntax is more complex when it has more asymmetries (e.g., German is SVO in main clauses, SOV in subordinate clauses, and requires SV inversion after an initial adverb; cf. English
Never have I seen such a mess
) and/or more overt or grammaticalized expressions for more semantic and/or pragmatic distinctions than another language. He makes no claims that these complexities come with more difficulty of production or processing, since he assumes that human cognition is capable of processing great degrees of what he calls
overspecification
. Linguists have long recognized that the phonemic inventories of smaller, isolated languages tend to two extremes: either very many or very few. The ones with very many phonemes have complex phonologies by definition, while the ones with very few phonemes exhibit their complexity in word length, as in the case of the name of Queen Lili‘uokalani born Lydia Lili‘u Loloku Walania Wewehi Kamaka‘eha, last monarch of Hawaii. McWhorter wants to extend the study of relative complexity into all parts of the language, understanding that the label
complexity
has no qualitative value.

At work here is the perhaps surprising factor of second-language learners. In languages with many phonemes, the phonetic space is crowded from the point of view of second-language learners and is therefore difficult to acquire. In languages with few phonemes, words tend to become long, which puts a load on working memory for second-language learners. Languages with significant numbers of second-language learners tend to have 20–50 phonemes, and English is exactly in the middle of that
range. What is at issue here is primarily the consonant inventory, with 26 of the 39 Standard English phonemes being consonants.
¹¹

Second-language learners also play a role in the structure of languages that have spread across an area by means of language shift. These languages also tend to be less complex in McWhorter's terms and therefore also more regular and transparent. When it came to the Altaic languages that successively spread across the Eurasian steppe, Nichols notes that they were well designed not just for translatability with other languages they were in contact with but also for second-language learners (Nichols 2010:191). So, among the many conditions to which languages are always catching up is: whether they are spoken by small, isolated communities, in which case the linguistic structures can become elaborated over time and not particularly transparent to potential second-language learners; or whether they are spoken by larger groups in contact with other groups and second-language learners, in which case they may become structurally more regular. Their lexicons will derive, for instance, ‘teach' from ‘learn' and ‘show' from ‘see' and ‘drop' from ‘fall' with the same morphological process, such as the English causative -en found in ‘blacken' and ‘lighten,' but which is inconsistently applied, as the teach/learn, show/see, drop/fall suppletives show.
¹²

Structural complexity does not correspond to conceptual complexity. In English, one can say
I am in the poorhouse
meaning ‘I don't have a lot of money'. One can also say
I am house poor
meaning ‘I have a large mortgage and associated monthly expenses I can hardly keep up with'. These two phrases would be counted as structurally less complex than the corresponding ways they could be expressed in any of the Romance languages, for instance. The two phrases are, however, not conceptually less complex in English, and understanding them requires quite a lot of inference and cultural knowledge. David Gil, one of the editors of
Language Complexity as an Evolving Variable
, contributes a chapter in his volume entitled “How much grammar does it take to sail a boat?” He is a specialist in Indonesian, and he offers examples of four different colloquial varieties of this language, all of which exhibit the structurally noncomplex features of:

1. no word-internal morphological structure;
   
2. no distinct syntactic categories; and
   
3. no distinct construction-specific rules of semantic interpretation.
   

He notes that: “Speakers of these different varieties range from westernized and upwardly mobile office workers in high-rise buildings in Jakarta, through shopkeepers and rice-farmers across the archipelago, all the way to New Guinea highlanders in penis gourds and grass skirts” (Gil 2009:28). His point is a fine inversion of the old technological complexity–linguistic complexity association. And on our most complex technology to date, namely our smart phones, the motto for our texting practices could be: keep it simple.

When the Spanish made their first settlements in California in 1769, some 250,000 Native Americans lived there. From what is available in the historical records, these peoples can be grouped into 21 known nationalities, or small nations, which can be
further separated into subnationalities, tribes, and tribelets to come to a total of 250 distinct groups. The concentration of languages associated with these groups in this area can be divided into six major phyla, and their number is the reason historic California is considered a residual zone. The relationships among the tribes, at least to outside observers, were described as one of intimate separatism. The arrival of the Europeans did not cause them to band together against the invaders, nor were they particularly interested in assimilating to the new and dominant culture. The California Gold Rush of 1848 marked the beginning of the end for many tribes, and the 1860s and 1870s were the years when the clashes – meaning massacres – between the Indians and the Whites reached a climax.

During this period, one tribe of hunter-gatherers, the Yahi in northern California, was reduced to a small group of people, which included a boy born around 1860. Forty-one years later, in 1911, this boy had grown into a man and was the sole survivor of his tribe.
¹³
Emaciated and without any clothing except a covered-wagon canvas he wore around his shoulders, he was discovered by Whites living in Oroville, California who took him to the sheriff. After attempts to speak with him in a variety of languages failed, Professor Thomas Waterman of the University of California was brought to town. Waterman was a student of Franz Boas, and he specialized in Native American languages. After some investigation, he came to understand that this man spoke a variety of the Yana language. He was called Ishi after the word for ‘man' in his language. The name
Yana
means ‘person, human being,' and over and over in English the tribal names for California Indians come from the answer to the European question “Who are you?” which prompts the usual Indian answer, “I am a person.” What else to answer? To the Native American, the question is a rude one, if not rudely meant. It was not the usual custom in Indian society to bandy one's name about, and certainly not to a stranger.

Ishi lasted five more years among the Whites, who treated him well, but he succumbed to tuberculosis 1914 and died of it in 1916. During his time among the Whites, he provided as much information as he could about his culture and language. He made a map of Yana territory with villages and trails, and heads of salmon runs. He demonstrated how he made his bows out of juniper wood and arrowheads out of stone. He was Edward Sapir's informant for a Yana grammar and Yana folk tales and lore. Like many Native American languages, Ishi's variety of Yana was polysynthetic and had male and female language, that is, separate forms for verb endings, pronouns, and demonstrative pronouns used by the men and the women.
¹⁴

The point of this story is that the last Stone Age man in North America spoke a fully elaborated language, and he brought it forward into a world that knew steam engines, telephones, automobiles, and airplanes. Whatever might have been the beginnings of language 100+ kya, anything resembling a primitive language became elaborated tens of thousands of years before the first Americans crossed the Behring Straits.