You are on page 1of 13

Do Pirahã speak Nean?

Yuri Tarnopolsky

The war of words is a metaphor that should be taken literally in linguistics.

In this essay I am omitting the history of the long war between the formal linguistics of
Noam Chomsky and his critics because the linguists know the story and the non-linguists
can turn to the well written article in The New Yorker: The Interpreter by John Colapinto
(April 16, 2007). The war remains as unfinished as the Iraq war, but the general state of
things is more or less clear: the Gullivers lose to the Lilliputs.

The article in The New Yorker is an excellent piece of journalism. It can be

complemented by many available on the web professional materials googlable as: “Dan
Everett” Piraha OR Pirahã. I have no say in a professional linguistic discussion, but
my point of view may be of interest for the seekers of new ideas.

John Colapinto actually traveled to the Pirahã tribe in Amazonia together with Dan
Everett, a linguist and an unordinary personality, who is an unrivaled speaker of the
exotic language of the tribe.

John Colapinto writes (p.130):

When I asked Everett if the Pirahã could say, in their language, "I saw the dog
that was down by the river get bitten by a snake," he said, "No. They would
have to say, “I saw the dog. The dog was at the beach. A snake bit the dog.”

I recognize in Everett’s transposition the language that I called Nean in my web


1. Pattern Theory and “Poverty of Stimulus” argument in linguistics (PoS)

2. Tikki Tikki Tembo: The Chemistry of Protolanguage (Tikki)

and others at complexity.

I am a chemist and not a linguist, although I am superficially familiar with the basic
structure of several very different languages. My interest in languages is part of a larger
program “a chemist’s view of the world” pursued on my website.

I see chemistry as one of the two most romantic sciences; linguistics is the other one.
There are many parallels between them, which I explore at my site. Thus, chemistry
uses an artificial language that compresses the rich three-dimensional world of
chemical structures into a strictly linear sequence of symbols, exactly as our language
does when we describe an elephant or the global warming. “The language of DNA” is
not a metaphor, but a standard term in molecular biology. Mark Baker, an eminent
linguist, entitled his book “The Atoms of Language.”

Here is just one example, which I never used before. One of the starting assumptions of
the entire theory of Noam Chomsky about universal grammar was the ability of
children to construct correct sentences never heard or corrected before (the so-called
poverty of stimulus argument). An average chemist during his or her life brings to
existence large numbers of chemical substances that never existed before, probably,
even in the entire solar system. Each gets a unique name that serves as a photo ID: you
can draw the molecule from the name and materialize it in the lab, if you wish.

I am a father of several dozens of never before known substances, but there is nothing
to be proud of: anybody can do that. Does it mean that I have some innate knowledge of
chemistry? Of course not. Chemistry had even denied a possibility to make some of my
molecular children before I actually did it.

I and the rather happy looking and attractive Pirahã Indians, whose photos can be seen
in The New Yorker and on the web—we all have an innate ability to understand the
world and acquire skills and knowledge. If the adult Pirahã cannot learn counting, it
means that they either do not need it or the teacher cannot find the right key. If I am not
mistaken, some of our very distinguished citizens could not master the Internet, either.
Inability to learn the new lessons of history is even more common, and for the same
reasons that Dan Everett attributes to the Pirahã: “This is new stuff and they don’t do
new stuff.”

Most normal people in this hemisphere are not only incapable of learning chemistry but
consider it the most incomprehensible and even repulsive area of knowledge. There are
plenty of other occupations, for example, making and counting money. Some of our
top national leaders occasionally fail even at their supposedly innate language skills,
instead relying on somber language of power—a language much more poor, color-blind,
and primitive than even the language of the Pirahã.

But what is Nean? In short, it is speaking in triplets or even in doublets. “I see dog.
Dog was beach. Snake bite dog.” This is a good example of Nean grammar. It does
not sound good, but all first contacts between different tribes and nations started with
learning a pidgin version of the other language. The study of pidgins is a separate
branch of linguistics. The fact of great importance is that we understand “I see dog”
and “Dog was beach,” especially, in context. It is probably obvious to any Amazonian
that snakes are common on beaches. He can explain to the crooked head: “Snake was
beach” or simply “snake beach! snake beach!” “There are many snakes at the beach”
sounds like “snake snake beach! snake beach snake snake!”

Nean, in my view, is not exactly a language, but the primeval natural grammar. That
was my first thesis of an outsider when I got interested in the origin of language some
years ago.

I suggested in PoS and Tikki that Nean is a grammar of such simplicity that it could
originate spontaneously by self-organization. This follows from my perception of

complex systems: any natural complex evolving system, such as life, society, culture,
technology, etc., starts spontaneously as a system of minimal complexity. Then it
grows, evolves, beefs up its complexity, and diverges. Only very simple things in very
simple systems can happen without the input from a deity, human, or alien. A group of
monkeys cannot type Hamlet, but they can certainly type by accident the name of
Hamlet. Names like Bush and Gore are even more—and equally—probable to pop up in
such experiments. No political allusion intended.

NOTE: At a closer look, to type “Bush” may require less physical energy than
to type “Gore.” “Gore” takes longer jumps from key to key. But this is arguable.
What matters is the approach to language as to a natural process. What takes
less energy is more probable.

On the contrary, if a system is created by human mind from some existing material, it
may not need simplicity. Bureaucratic systems are the best example. It has been noticed
that our appliances are getting more complex and less reliable. The natural history of
bureaucracy as pattern, however, also had to start with something very simple, such as
just counting or policing. Moreover, counting also had to start with counting to two,
further proceeding to tree, five, and ten or twelve.

In order to advance in mathematics further, you need a mathematician. The same with
language: you need a speaker and a writer. But you need Demosthenes or Cicero only in
a developed society. They are not expected to be good hunters in the jungle.

My second thesis was that the system consisting of the speakers and the social
environment maintains the state of homeostasis, which, probably, is the case with the
Pirahã tribe. There is no natural language separate from nature. Language is the device
for maintaining the homeostasis of the tribe. If it evolves, it is because the homeostasis
has been disturbed. If not, people can manage with what they have.

My third thesis regarding language was that we have two very different kinds of
languages. One (Language1 or hypolanguage) is the language spoken today in a much

more advanced form only by some uneducated and illiterate, but otherwise completely
normal, gentle, skilled, and caring people, albeit with limited experience, which we can
find only in particular areas and social strata of the world, even in quite civilized
countries. It is also the language of children before some age and level of schooling. It is
also the language of new immigrants who are not as articulate as Noam Chomsky, but
their children may become one. It consists of short segments of speech, often repetitive
and overlapping. It is practically free of embedding and recursion. It is simple because it
is difficult. For the same reason Hamlet is more difficult than “Hamlet” and “Hamlet” is
more difficult than “Gore.” But the cardinal fact is that we understand Language 1 and
people can successfully communicate in it, up to a point.

The second language (Language 2 or hyperlanguage), that of long composite phrases

with many subordinate clauses, is the result of advanced evolution. It is a bulky human
artifice, like HMS Queen Mary II. It was gradually created, by trial, error, and mutation,
in order to represent the growing complexity of civilization, including complex ideas and
constantly shifting homeostasis. It is the language of science, literature, and bureaucracy.
It has nothing to do with the origin and essence of the language that all humanity speaks,
but everything to do with culture, society, economics, and maybe a desire to show off.
On the negative side, it is also the truly barbaric language of the fine print and the US
Tax Code—the language so alien, that we pay big money to professional translators.

With all the fuss around the Pirahã, I feel a need to find some new non-professional
arguments, just to convince myself.

All Soviet college students in my time were mobilized to spend at least a month doing
some agricultural work—a kind of forced labor. While doing time in the Russian
countryside, I witnessed the aboriginal Russian and Ukrainian language, which used
probably one percent of the grammatical resources (and sometimes 200% of the foul
vocabulary for the male speakers). It was not a local dialect or slang, but the natural
speech of some people of mostly older generation, sometimes almost illiterate, and born
before the introduction of the comprehensive system of education that the Communists
had among their undeniable achievements.

I of course do not have any examples. I remember, however, that in spite of its simplicity,
the language was by no means handicapped. It was very expressive, for which there are
incredibly rich lexical and morphological resources in both Russian and Ukrainian. It was
in homeostasis with environment. It served its purpose and function. I understood the
speakers and they understood me.

Stimulated by the story of Pirahã, I decided to look at the records of Russian folklore that
could open an audio channel to the past..

There is a Russian site dedicated to Russian literature and folklore:

Here is an example of folk poetry collected in Russia in the middle of the eighteenth
century by Kirsha Danilov (in my translation).

"Гой еси вы, князи и бояра Hey, you, princes and boyars,
И могучие богатыри! And mighty heroes!
Все вы в Киеве переженены, You are all married in Kiev,
Только я, Владимер-князь, холост хожу, Only I, prince Valdimir, remain single,
А и холост я хожу, неженат гуляю, I remain single, walk unwedded,
А кто мне-ка знает сопротивницу, And who knows a mate for me,
Сопротивницу знает, красну девицу: Knows a mate, a handsome girl:
Как бы та была девица станом статна, Be that girl slender of figure,
Станом бы статна и умом свершна, Slender of figure, clever in the head,
Ее белое лицо как бы белой снег, Her white face like white snow,
И ягодицы как бы маков цвет, The cheeks like the scarlet poppy,
А и черныя брови как соболи, Black brows like sables,
А и ясныя очи как бы у сокола". And the lucent eyes like the falcon’s.

The only possessive pronoun is highlighted. Otherwise, there are no obvious

grammatical connections between the lines. There is no embedding. The short segments
are brought together by the repetitions that look like the “pre-existing condition” of
haplology, which I discuss in Tikki:

The lines look ready for haplology contraction:

And who knows a mate for me, Knows a mate, a handsome girl: Be the girl
slender of figure, Slender of figure, clever in the head, Her white face like white

They can contract into the single phrase:

And who knows a handsome girl with a slender figure, clever head, and a snow
white face, who could be a mate for me?

The folklore is not what we can call natural speech. It is a one-way communication and a
product of long evolution, polished by centuries. It can be seen as an intermediate step
between Language 1 and Language 2: “Language 1.5.” The poems in the collection of
Kirsha Danilov relate to the beginning of Russian Christian statehood, around 1000 A.D.,
as well as to more recent times.

Here is another example:

"А и ласково сонцо, ты Владимер- You are indeed a gentle sun, prince
князь! Vladimir!
Не нада мне твоя золота казна, I don’t need your golden treasure,
Не нада три ста жеребцов I don’t need three hundred stallions,
И не нада могучия богатыри, I don’t need mighty heroes,
А и только пожалуй одново мне Only give me one lad,
Как бы молода Екима Ивановича, Like the young Ekim Ivanovich
Которой служит Алешки Поповичу". Who works for Alesha Popovich.

In the samples of old Russian poetry, the word которой (kotoroy) is the relative

pronoun who with a male gender ending. Used as a relative pronoun, it can be found
elsewhere in the collection of Kirsha Danilov, but it is rare: 8 per 8000 words (0.001).
Ten more are interrogative pronouns. In an at random selected segment of Leo Tolstoy’s
Childhood, the frequency of kotor- plus case and gender ending, used as relative
pronoun is 17 per 2512 words (<0.01). Tolstoy, however, uses all other available subject
pronouns and high level of recursion (still negligible as compared with the American Tax

The two last lines of the above example could be put differently, in the tradition of “pre-

Как бы молода Екима Ивановича, Like the young Ekim Ivanovich,

Еким Ивановичa при Алешкe Ekim Ivanovich with/at Alesha Popovich.

I can even imagine a longer pathway to a subordinate clause with “kotory” (which, that,
who). Kotory is also the interrogative pronoun (who? which? what?). The pathway is
somewhat reminiscent of the “knock-knock” jokes.

Как бы молода Екима Ивановича, Like the young Ekim Ivanovich,

Екима Ивановича которого? Ekim Ivanovich who?
Который при Алешкe Поповичe." Who is with Alesha Popovich

Apparently, there was a lost in time person who applied haplology and invented the
pronoun out of something available for another purpose, like a primitive generalized
interrogative pronoun, or a deictic particle, like the mysterious ‫( כי‬ki) of the Biblical

Hebrew (in ‫ כי־טוב‬, ki-tov of Genesis and thousands of other occurrences), which can
stand for practically anything, from yea, indeed to except that, see Selected References.
Francois Jacob called this tendency of biological evolution to use and combine existing systems in a new
way bricolage , tinkering. Evolution of language could appropriate the same style, probably, because there
is only one evolution of everything.

My point is that the Nean grammar, in all its natural primitiveness, opens a way to a
further gradual evolution toward modern recursive language, which we, however, try to
avoid in everyday communication. The dense recursivity is intended for the written
language, which does not impose a heavy burden on memory. The written text keeps a
large chunk of the narrative fully in view, or, to put it differently, in ROM, Read Only
Memory, which, alas, is too short in the natural, unaided mind.

Regarding Pirahã, the controversy comes not from its peculiarity but from the theory that
was built initially on Language 2 but tried on Language 1 or 1.5.

I believe that the painful Frankensteinian story of formal linguistics, re-tailored, re-cut,
and re-patched many times over (but so beneficial for scores of graduate and
postgraduate students of humanities all over the world), would come to an end long ago

if Noam Chomsky started with simple languages of pre-literate people instead of the non-
existing infinite productivity of grammar under the pens of sophisticated writers. It looks
like John Colapinto also noted that.

My prediction is that with more time the Pirahã, the immigrants from our past, who
suspect the outside traders of cheating, will find a way to control their trade, probably, by
letting their children learn math and Portuguese. In the immigrant families it is the
children who serve as an interface between the old folks and new reality.
Until then, it is just a tribe whose homeostasis has been miraculously preserved, as Dan
Everett’s remark in his main paper testifies: “The Pirahã are some of the brightest,
pleasantest, most fun-loving people that I know.”

If Russian is a language, so is Pirahã. But probably not for the reasons of the formal

As for the formal linguistics, a similar story happened with chemistry: it was prevented
for a long time from the true understanding of nature by the dogmatic theory of
phlogiston, the hypothetical substance of flammability with zero or negative weight:
sulfur burns and disappears because it consists of phlogiston. During the entire eighteenth
century large number of contradicting treatises had been written in order to make some
sense of the theory—quite like with universal grammar—before Antoin Lavoisier
knocked it off like a dead cockroach. He then lost his head in the French Revolution, but
this is a different story.

Quite unexpectedly, the Pirahã story has made me appreciate the historic significance of
Noam Chomsky as a public figure on the world scale. He has been keeping linguistics in
the focus of public curiosity and attention. As result, the linguistic trade today is alive,
growing, and hot. Naturally, mosquitoes from the young swarm have come to bother the
master and grow through the ranks on his blood. Similarly, as a stinging gadfly himself,
he has been driving the pachydermous American foreign policy into the focus of world
attention, although the animal was quite capable to stay there on its own. Whether he is

right or wrong is of secondary importance as compared with the importance of his

dogged search for truth. Besides, can a person of Chomsky’s intellect be completely

John Colapinto got his story right. In politics follow the money, in private life cherchez
la femme, and in linguistics listen to the horse’s mouth.


The frequency of the use of Russian relative pronoun который (kotory : which,

that, who) in all its number (2), gender (3), and case (6) forms (total of 36 combinations)
gives some idea about the level of recursion.

This is what I found, using online sources of Russian texts.

Frequency of relative pronoun kotor+ending (which, that, who) in Russian

literature since ancient times, from selected sources
Source Time Frequency Note
(sample size)
1 Byliny (epic stories Probably, most 0 Rich poetic and
and songs) ancient (over 7,000 magic content
words) see
Russian Appendix 2
2 Skazki (fairy tales) Open question; 0.0002 Intended for
no canon exists (7,800 w) children, very
3 Autobiography of 17th century, 0.0005 Coarse everyday
Arch-priest Avvakum (20,000 w) language of the
(see Web) time
4 Collection of Kirsha Recorded in 0.001 References to
Danilov 18th century historic events
(8,000 w)
after 1000 A.D.
5 Leo Tolstoy, 1852 0.0093 Highly recursive,
Childhood , story (14285 w) emotional style

Childhood , story emotional style

6 Fedor Dostoevsky 1866 0.002 Feverish,
Crime and (39000 w) unpolished, close
Punishment, to common
novel, Part 2. language
7 Anton Chekhov, 1896 0.004 Refined, lucid
My Life, story (29,300 w) style, monologue
8 Leo Tolstoy, 1873-1877 0.007 Epic Tolstoy’s
Anna Karenina, (14,500 w) style
9 Leo Tolstoy, 1899 0.0093 More animated
Resurrection, novel (10,745 w) late style
10 Anton Chekhov, 1904 0.002 Mostly
Cherry Orchard, play (13,000 w) dialogue

I do not attribute any scientific validity to my little investigation, but it seems to illustrate
how recursion (better to say, syntactic complexity of language) evolves over time and
depends on the individual style at advanced cultural stages. I used limited sources and
only a part of the Byliny and Skazki texts.

APPENDIX 2 (April 30, 2007)

The zero occurrence of kotory in Russian byliny (plural of bylina) seemed suspicious to
me. Many byliny in the large Russian collection were recorded in recent times, as late as
1956 and 1980. The languages of the teller and the listener must be at all times more or
less compatible. Therefore the language of oral folklore could also evolve: the teller is
also a speaker. When I repeated my amateurish investigation on a much larger collection
of byliny texts (over 55,000 words) I found the occurrence of kotor to be around 0.001. It
was, however, clustered only in selected few byliny and not distributed uniformly. The
majority of the byliny go without a single kotor . This confirms in my eyes that the
recursion and, more generally, complexity of language can be the property of culture,
location, and personality of the speaker, instead of the language itself.

I can also formulate a hypothesis, without any evidence at this time, that the language,
following the evolution of culture and discovery of the world, first increases its
complexity to fit the complexity of the world, but later begins to shrink under the kinetic
constraints, i.e., the need to say a lot in a short time. There is always a gap between
written and spoken languages. But even the written language—like that of the airport
books—flattens and crumbles under a new cultural constrain of dumbing everything
down in the competition for the customer base. Commerce takes over culture.

But what about old sources in English language?

A cursory look at Chaucer’s The Canterbury Tales (1400) shows the modern
sophisticated grammar: no wonder for a heir of Latin literature, in contrast to Russian

And al above ther lay a gay sautrie,

On which he made a-nyghtes melodie
So swetely that all the chambre rong;
And angelus ad virginem he song;

The original, now dark, Middle English language of Chaucer was preserved by writing
and printing, while the Russian language of the byliny, completely understandable today,
was, probably, rolled smooth like the beach pebbles by the tides of time.

It would be interesting to look at other ancient and minimally processed spoken sources .

The Finnish joka (which, that, who) exists in about 20 case and number forms, luckily, not complicated by
the gender, absent in Finnish. It seems to be sparsely present in Kalevala.

Otti ruunansa omansa, He took his horse,

jonka turpa tulta iski . . . whose mouth (face) striked with fire . . .

Jonka is Genitive of joka. In its unchanged form joka also means each, all. I wonder how much the
original records of Kalevala were edited by Zakarias Topelius and Elias Lönnrot. The Lönnrot‘s archive is
preserved in Finland.

Notably, the Pirahã language does not have each and all.

APPENDIX 3 (June 7, 2007)

To summarize this essay in a few words, we do not need to go to Amazonia in search of

primitive languages: they could be found within a hundred mile radius from a large city,
probably, in the closest prison or on a construction site, and they could sound very much
like English, Russian, or, probably, even French, depending on where you are. Benjamin
Whorf captured the mirror image of reality: the way of life, both as a snapshot and as a

history, shapes the language, rather than the opposite. But it does not mean he was
completely wrong: you need only to flip the image.

But, as I said, Noam Chomsky cannot be completely wrong. That would be against the
laws of nature. Where is he right? Instead of a long answer I would refer to the absolutely
unique book (some justly say stunning, I am tempted to say divine) by Eric Kandel In
search of Memory.(W.W. Norton, 2006) . It is about innate mechanisms, among many other
fundamental things.

I believe that the experimental kinetic linguistics that measures the grammatical
complexity and compression of language of social groups (by age, too) and plots it along
the axis of cultural intensity and complexity of life in the group would illuminate
linguistics in the same way Eric Kandel illuminated the science of mind.


1. Dan Everett’s main paper:


2. The detailed paper of Andrew Nevins, David Pesetsky, and Cilene Rodrigues, which criticizes the
conclusions of Dan Everett, can be accessed through (click on the title)
or directly downloaded through

3. A sample of Piraha text:

4. Peter Gordon’s article in Science is available online through EBSCOhost Research Databases:
Numerical Cognition Without Words: Evidence from Amazonia., By: Gordon, Peter, Science, 00368075,
10/15/2004, Vol. 306, Issue 5695 , pp. 496-499 Database: Academic Search Premier. Of course, you
can drive to a big library, if you have nothing better to do.

On Hebrew particle ‫( כי‬ki):

5. Carl M. Follingstad. Deictic Viewpoint in Biblical Hebrew Text: A Syntagmatic and Paradigmatic
Analysis of the Particle ki . Dallas: SIL International, 2001.

6. Review:

Nicolai Winter-Nelson, Deictic Viewpoint in Biblical Hebrew Test: A Syntagmatic and

Paradigmatic Analysis of the Particle (ki), Journal of the Evangelical Theological Society, Dec.
2002 . or

April 15, 2007 email

Updated: June 7, 2007