The Pirahã Controversy — Part 3

Mar 29, 2012 by

In the previous post, I outlined the arguments in the Pirahã controversy. Here, let’s consider more closely one of the most crucial issue of this controversy: whether Pirahã allows recursion. (An interested reader is referred to the delightfully humourous and culinarily informed post by Daniel Harbour on this topic.) Numerals and other alleged linguistic gaps in Pirahã will be discussed in a separate post (or posts).

When it comes to defining recursion, two distinct concepts that may be understood under the term “recursion” need to be distinguished: putting something inside of something which is then put inside of something else, and putting objects of the same type inside each other. Daniel Harbour appetizingly describes the two types of recursion as “dumpling-borsht recursion” (filling inside dumpling inside borscht) and “turducken recursion” (chicken inside duck inside turkey). The really crucial type of recursion is the latter type, where it is objects of the same type that are embedded inside one another, like Matreshka dolls. Dan Everett himself shares this definition of recursion: “putting one [multiword] phrase inside another of the same type …, e.g., noun phrases in noun phrases, sentences in sentences, etc.” (Everett 2005: 622). Thus, constructions involving embedding include, among others, recursive possession (noun phrase inside a noun phrase) and clausal objects (a clause inside a clause). The following English examples illustrate the relevant structures, which are bracketed in the examples below:


a. Recursive possession:           [Mary’s brother’s] canoe has a hole.

b. Clausal object:                           John knows [how to make an arrow].

As it turns out, Everett is right in that the Pirahã language does not have the constructions corresponding precisely to those in the English examples in (1). But this is not the whole story, as Pirahã has variations of these constructions that are found in many other (unrelated) languages.

To start with the recursive possession, while a prenominal possessor is possible in Pirahã, a possessor noun phrase may not itself contain a possessor:



xipoógi             hoáoíi               hi         xaagá.

Xipoogi            shotgun            3rd        be

‘That is Xopoogi’s shotgun.’ (Everett 1986: 205)


*kó’oí  hoagie  kai                   gáihií    ’íga.

Ko’oi   son       daughter           that      true

‘That is Ko’oi’s son’s daughter.’ (Everett 2005: 630)

But this ban on recursive possession not something peculiar to Pirahã or other similarly “exotic” languages. It is also found in such familiar languages as German:



Hans-ens         Auto

Hans-poss       car

‘Hans’ car’


*Hans-ens       Auto-s                         Motor

Hans-poss       car-poss          motor

‘Hans’ car’s motor’

It has been proposed that this prohibition against possessor recursion in German has to do not with the impossibility of embedding but with limitation of the genitive case; the same effect is found in other genitive environments in German, such as the direct objects of particular verbs and prepositions. Crucially for our discussion of Pirahã, the fact that German appears to show the very same restriction suggests that “whatever syntactic switch turns off prenominal possessor recursion in German is also at work in Pirahã” (NP&R 2009: 368). Furthermore, it is pretty obvious that “the culture shared by most German speakers is more similar to that of most English speakers than either English-speaking or German-speaking cultures are to the culture of the Pirahã” (NP&R 2009: 368), so the explanation for the lack of recursive possession cannot be formulated in cultural terms.

The arguments based on clausal embedding is similar: the ways in which clausal objects are structured in Pirahã are similar to those found in other languages, which do not share the cultural peculiarities of Pirahã. For example, clausal objects in Pirahã must be contain a special morpheme -sai which said to make the object clause more nominal (e.g., grammatical notions such as tense, aspect and agreement cannot be expressed by such nominalized clause):


hi         ob-áaxáí                      [kahaí   kai-sai]

3rd        see/know-intns          arrow   make-nmlz

‘He really knows how to make arrows.’ (Everett 1986: 263)

However, using nominalized constructions for clausal objects is not just a common trait among Amazonian languages, but it is found in many other languages including Quechua, Turkish, Inuktitut, and Adyghe.

Moreover, Everett notes that in Pirahã clausal objects, as in (4), follow the verb while nominal objects precede the verb; for example, the nominal object kahaí ‘arrow’ precedes the (nominalized) verb kai-sai ‘make’, but the bracketed clausal object follows its verb obáaxáí ‘really know’. However, this combination of OV order with nominal objects and VO order with clausal objects is also far from being rare cross-linguistically; a sampling of languages with this property includes German (illustrated below), Hindi and Wappo (an extinct Yukian language, once spoken in California). In the German examples below, the object is bracketed and the relevant verb is boldfaced.


a. OV (nominal object):

Hans    hat       [die      Kinder]            gesehen.

Hans    has       the       children            seen

‘Hans has seen the children.’

b. VO (clausal object):

Hans    sagte, [dass     er         die       Kinder             gesehen            hat].

Hans    said      that      he        the       children            seen                 has

‘Hans said that he has seen the children.’

Upon an examination of various properties of Pirahã that Everett claims to be specially constrained by the speakers’ culture, NP&R’s (2009: 359) conclude that

“if speakers acquire the same types of languages whether their home is a German city, a village in the Caucasus, or the banks of the Maici River in Amazonas, Brazil, we have discovered just the kind of disassociation between language and culture that sheds light on the nature and structure of UG”


And while this conclusion runs contrary to Everett’s view of Pirahã as being surprising and unique among human languages, this result is extraordinary in its own way and “not at all mundane, in the end” (NP&R 2009: 359).


Related Posts

Subscribe For Updates

We would love to have you back on Languages Of The World in the future. If you would like to receive updates of our newest posts, feel free to do so using any of your favorite methods below:

  • Serge Minor

    Wouldn’t it be better to use the Perfect tense of ‘sagen’ in (5b) to have a minimal pair with (5a)? When the verb is in the Imperfect form in German the order seems to be obligatorily VO even with nominal objects due to verb-second effects, if I’m not mistaken.

  • Since Everett has disclaimed the analyses (not the examples) in Everett 1986, it’s shaky to draw conclusions from them.

  • Since Everett has disclaimed the analyses (not the examples) in Everett 1986, it’s shaky to draw conclusions from them.

    • Other people, like Uli Sauerland, have done work that confirm the glossing of these examples (Uli’s work in particular is about -sai).

  • Ivan Derzhanski

    Both the report that (2b) is ungrammatical in Pirahã and the retort that (3b) is likewise ungrammatical in German would be more enlightening if we knew whether Pirahã has an alternative way of saying ‘Ko’oi’s son’s daughter’.  You can say ‘Hans’ car’s motor’ in German all right, simply by using another possessive construction: _das Motor von Hansens Auto_.

    • There is no other possessive construction reported in Everett’s fieldwork, as far as I know. It is my understanding that the only way to render ‘Ko’oi’s son’s daughter’ in Pirahã is by using several utterances, something along the lines of “Ko’oi has a son. The son’s daughter…”

      • Ivan Derzhanski

        But that makes Pirahã truly unlike German (or most other familiar languages), doesn’t it?  It may well be true that it is the same syntactic switch that makes (2b) and (3b) ungrammatical, but the fact that Pirahã offers no means whatever for expressing recursive possession does constitute a claim to uncommonness.

        • Or else it is something that escaped Everett’s attention… him trying to prove the uncommonness of Piraha, it wouldn’t surprise me…

  • Congrats, Asya. We need blogs such as yours more than ever, to bridge the enormous gap between academia and the general public (since the press does such a poor job at that).  A deeper issue that seems to be overlooked is the fact that Dan was a Christian missionary and missionaries are, obviously, engaged in triggering cultural change, whether that is as benign as teaching math (, or as fundamentalist as expelling “spirits” from the Indians’ lives ( So, even if the Pirahã had such a unique culture (I think all cultures are unique, but the uniqueness of Pirahã has been blown out of proportion), that uniqueness may have been tainted by the contact with missionaries.  The “math classes” excuse (which seems to have come almost as an afterthought to justify why certain experiments didn’t corroborate the Everetts’ theory) is just an example of how the Pirahã may not be considered as “untouched” by western civilization as Dan wants us to think (in part because of his (and his family’s) own work there).

    It is not surprising that Dan’s theory is particularly popular among those who have no familiarity with indigenous South American languages, such as Mr. Oliveira, a Brazilian professor quoted in the NYT piece.  Among South Americanists, most of his claims of Pirahã uniqueness fail to convince. Take the numerals situation, for instance. The overwhelming majority of lowland South American languages simply lack a productive class of numerals. Most have numbers only up to three or four. And even these few tend to be rather unstable diachronically. Even for a close-knit family such as Jê (whose proto-language I’ve been reconstructing), it is impossible to reconstruct anything beyond “one” (which also means “alone”). And consider the claim that Pirahã originally lacked pronouns, having borrowed them from a Tupi-Guarani language ( It is pure speculation, of course, since nothing is known about the past history of the language (and that will remain so, since the only known relative of Pirahã, Mura, is now extinct and was very poorly documented). It requires a leap of faith, assuming (with no proof whatsoever) that the language originally lacked pronouns and that that gap would have been filled by contact with Tupi-Guarani languages (even though Pirahã lacks any lexical evidence of contact with Tupi-Guarani languages). It is a rather clear example of shaky methodology being used to corroborate the author’s preconceptions on the “primitive” nature of the language.

    • Thank you for your interesting comments and link, Eduardo! I am hoping to be able to blog more on the Piraha numerals this week, so stay tuned. And thank you for your kind words about the blog!

  • Marius Svenkerud

    It’s not actually true that German prohibits (3b). Sure, the precise phrase ‘Hansens Autos Motor’ is not possible, but that’s because of a perfectly irrelevant restriction on the use of ‘Autos’ (and, in fact, most other words). Here are some examples from the Web:

    Jennifers Vaters Frau                         ‘Jennifer’s father’s wife’
    Jans Mutters Wunsch                       ‘Jan’s mother’s wish’
    Svens Vaters Geburtstag                   ‘Sven’s father’s birthday’
    Sandras Mutters Geburtstag             ‘Sandra’s mother’s birthday’
    Sebastians Vaters Geburtstagsfeier ‘Sebastian’s father’s birthday party’
    Stefans Vaters Haus                           ‘Stefan’s father’s house’
    Stefans Vaters Motto                         ‘Stefan’s father’s motto’

    Also interesting are the description ‘Vaters Vaters Vaters Mutters Vaters Mutters Vaters Mutters Vaters Mutters Mutters Mutters Vaters Mutters Vater’ and the sentence ‘Luke ich bin dein [sic] vaters mutters onkel dessen bruders wg mitbewohner’.

    • Thank you. It seems that “Vaters” and “Mutters” is exceptional in this respect…