Poverty of the stimulus

The Poverty -of-the -stimulus argument ( POSA ) (Eng. about " argument from poverty of the stimulus " ) states that much too complex and too diverse from what is in the human mind, is, as it may have occurred ( the course of life ) from the outside inside. The argument is often used by nativists in support of its position.

Nativism

Historically, nativism is the doctrine of innate ideas: Our ideas are in our minds from birth. Modern nativists see the seat of the " innate ideas " in the genetic makeup. But what exactly is innate? The ideas are not in the literal sense in the mind of the newborn. The appearance of the ideas in the mind is dependent on certain events or processes. Leibniz compared the mind with a block of marble. In the same way as the chisel brings the figure in marble to the fore, the events bring ideas to the fore. What is in our mind, is never enters into it, it was from the beginning it.

Nativists justify their attitude usually not by the positive proof that an idea is innate. One such positive evidence is indeed difficult to establish. The justification of nativism is the denial of empiricism. Particular importance is attached here to the Poverty of stimulus argument ( POSA ) to. Although the basic idea of ​​POSA dates back to ancient times, the term goes back to the linguist Noam Chomsky.

Chomsky POSA in particular

Noam Chomsky is one of the most prominent nativists of the present. In the area of ​​language is his POSA that the structure of language essentially can not be learned by data coming from the outside of an unprepared organism. ("[ The] narrowly limited extent of the available data ... leaves little hope that much of the structure of language can be learned by organism to INITIALLY uninformed as to its general character", p.58 ).

The basic structure of the argument is as follows:

• Therefore, people must have an innate and language-specific mechanism which includes knowledge of grammar.

Empiricists (such BBF Skinner ) argue, however, that there are general learning mechanisms are sufficient to tackle any complex task. Nativist turn, counter that the input in the course of life is too low. The general learning mechanisms were not sufficient to produce what we have to know in our minds. Chomsky concludes that there must be specific, language-specific mechanisms to explain the acquisition of language.

Basis: the universal grammar

Chomsky's starting point is the " amazing fact " that at the age of about eight years, nearly every child speaks the language of his speech community fluently. This succeeds children apparently without formal lessons and after they were previously faced with only a very small sample of sentences. In the early versions of his theory Chomsky children considered as de facto linguists who make hypotheses about the syntax of a language because of the input. To avoid endless " to grope in the dark," the child helps an innate universal grammar when setting up the hypotheses. The universal grammar was actually as a system of principles, conditions and rules understood by Chomsky (p. 29), which are present in all human languages ​​and to some extent, account for the essence of language.

Chomsky gave this approach later on in favor of the principles and parameters approach. That the universal grammar contains rules will be denied by him ( p 388 ). The universal grammar is rather to be compared to a kind of box with switches. The linguistic input that a child hears, causes certain switches are placed in the one direction or another. A "switch" is, for example, the decision whether you can omit a pronominal subject in a language or not (pro- drop parameter). In Italian, for example, you can omit the subject ( " Sono Italiano " ), in English and German can not ("I am German "). The knowledge that there are zero -subject and non-null -subject languages ​​is innate, and the decision coincides with the first language input; the switch is actuated in one direction or the other. With just a little input the child is in language development a major step forward, since there is a specific, innate mechanism that accelerates learning.

Three variants of the POSA

In support of these two aspects of universal grammar (congenital and language-specific) Chomsky draws on the Poverty of stimulus argument ( POSA ). According to Cowie, the POSA Chomsky is used in three variants.

The a posteriori POSA

This variant of the POSA is empirically basically: Since language can not be learned because of the available inputs, the principles of universal grammar must be innate. In response to a criticism of Hilary Putnam in Universal Grammar Chomsky formulated this form of POSA. Putnam was referring to Chomsky's finding that the mastery of the language is independent of the intelligence quotient of the speaker. Putnam replied that this only proves that any normal adult can learn what any normal adult learning. Of course, the " innate " human intellectual abilities are important for language learning. All in all missing Putnam evidence of a specific and innate ability to learn the language.

Chomsky refers to Putnam's finding that someone who is general learning mechanisms used will always use the simplest hypothesis. Chomsky brings an example to demonstrate that children do not just always use the simplest hypothesis in language learning. Suppose a child often hear phrases like these:

• " Ali is happy "
• " Is Ali happy?"

The child would have, initially following hypothesis for the use of general learning mechanisms:

Soon, however, the child will hear phrases like these:

• " The man who is happy, singing"

The child would have using H1 transform this sentence as:

• " Is the man happy, singing? "

The child should now be faced with negative feedback of the speech community, and then only form the following hypothesis:

This rule leads to the following question correctly:

• " Sing to the man who is happy?"

Chomsky argues as follows:

• The linguistic input is too low to allow the child might reject due to its H1.
• No child makes mistakes as in " Is the man happy, singing? ".

He concludes that no child of H1 per follows. Therefore, it also need not input that could refute H1. If the child acquires language using general learning mechanisms, there must be - according to Putnam's statement that the simplest hypothesis is to be preferred - initially show a preference for H1. Since this is not the case, there must be a language-specific learning mechanism give ( Universal Grammar ), which contains a rule like this: " Construct a structure -dependent rule and ignore structure - independent rules."

Chomsky's argument based on the fact that he assumes that H1 is easier than H2. H2 requires a syntactic analysis. H1 is based solely on the observation H2 refers to non- observable. This, however, Chomsky contradicts his own statement ( in the criticism of structuralism ) that grammatical hypotheses that refer only to observable, less easy and less elegant as such are hypotheses that relate to non- observable. According to his own testimony would have general learning mechanisms so H2 prefer.

Chomsky's assessment that structure-dependent rules are simpler than structurally independent, assumes that syntactic properties are related only through many intermediate steps with the linguistic experience of the language learner. Syntactic categories would therefore not be learned.

However, this assessment contradict the empirical findings. Saffran, Aslin and Newport could, for example, show that eight -month-old children after only two minutes to be able to distinguish in an artificial language between words and non - words, and that they thereby appear to use statistical regularities in the material. The only logical conclusion from this ( and many other ) experiments is that children can learn syntactic categories, by resorting only to general learning mechanisms.

Even Chomsky's statement that the input would be too low to enable children to reject H1 in a reasonable time, must be questioned. So found Pullum and colleagues in an analysis of the text base of the Wall Street Journal among the top 500 questions many that would refute a structure - independent rule. The same is true for an analysis of Oscar Wilde's The Importance of Being Earnest. Now, the Wall Street Journal and Oscar Wilde not the usual linguistic input for a child. However, should Pullums results on what children hear in the first years of life may indeed be transferable.

The POSA as a logical problem

This variant of the argument relies less on empirically testable assertions, but rather that the data in principle (a priori) may not be sufficient to enable the acquisition of grammatical rules. The language learners never hear ungrammatical sentences as " counter-examples " (negative evidence, negative evidence ). No competent speaker gives a child a list of wrong sentences and the notion that these are to be avoided. So stay the child to formulate only, even ungrammatical sentences and to have it rectified by the parents. But that almost never happen, parents were usually the ungrammatical utterances of their children. In addition, there are an infinite number of well-formed sentences that never hear the language learner. Therefore, the learners from the non- emergence of a sentence in the input can not conclude that this was against the rules. Consequently, language-specific knowledge must be innate.

However, the logic of this form of POSA is vulnerable ( p 215). If the absence of negative evidence in the input is sufficient to postulate an innate and specific mechanism, this could be also entered into many other areas. Imagine a person who learns what is a goulash. Almost every person obtained in the course of living a " culinary expertise " that allows him to distinguish between various forms of food and food of non- edible. No one but the person informed that tacos, pizzas, steaks and of course stones, dogs and clouds are not a goulash. The person is always confronted only with positive evidence of goulash. Despite the absence of counter- examples, all the people come to the view that a stew is a stew and nothing else. But no one would suspect behind an innate and dine specific mechanism.

The assertion is not true, there would be no negative evidence. For example, the fact that we call Hamburg "Hamburg " and not " goulash ", certainly a form of evidence that strengthens the idea of the speaker of " goulash ".

Above all, one should distinguish between " data" and " evidence ". Data are facts as they present themselves to the experience. Documents are data that are used for the confirmation or refutation of a theory. Data may be both negative and positive documents. Thus, a child does not necessarily require explicit corrections to know the falsity of a grammatical hypothesis can. Negative evidence can therefore feed from three sources:

Negative data as negative evidence

Nativists refer almost exclusively to a study if they claim children erführen virtually no negative feedback in language development. Brown and Hanlon had three mother - infant pairs were observed. The explicit confirmation or rejection of the mother regarding the spoken utterances of the child did not correlate accordingly with the well-formedness of what the child said. This seemed to confirm that negative feedback ( and feedback at all) has no influence on language development.

Later research, however, refuted what Brown and Hanlon had inferred from their data. Hirsh - Pasek, Treiman and Scheider man showed up about that mothers of children aged two ungrammatical sentences that their children much more frequently repeated ( and thereby corrected ) as the correct sentences. Hirsh - Pasek and colleagues conclude that the environment of the child is full of subtle references to the correctness of the statements made by the child. Further research clearly showed that the feedback that children receive the correct sentences, different from the is that children receive on incorrect sentences. Moerk was also able to show, in a reanalysis of the original data by Brown and Hanlon that even Brown's own records contain a wealth of korrektivem feedback.

Accordingly, it is simply wrong that parents do not correct the ungrammatical utterances of their children. That parents do not correct some of the ungrammatical utterances of the child, according to Demetras and other only a problem if one assumes, the child must master all at once the whole system of grammar. Children must not, as Marcus assumed to repeat the same sentence over and over again, to get enough feedback so that they can consider a rule. It is sufficient that they express sentences that are formed according to a certain rule, and to get feedback.

Positive data as negative evidence

A hypothesis can be refuted if one experiences positive data. For example, an English-speaking child that sentences like

• " The boy wants a curry" and
• "Dad wants a beer"

Listen, find the rule (1 ) confirms that the verb stem a -s must be attached forever. In terms of a ( by Steven Pinker adopted ) collecting restrictions now, one could imagine that the child is an arbitrary constraint ( 2) invents this rule, eg that this only applies if the subject is alive. This hypothesis of a limitation of the rule would be refuted once the child a sentence like

• "The curry tastes good "

Hears. Using this for the rule ( 1) positive data, the child may ultimately correct his over generalized rule (2).

The absence of data as a negative specimen

The language learner is to be regarded as an active hypothesis tester. If he finds that in his linguistic environment never a sentence is uttered, which would have to be a hypothesis according possible, he will reject the hypothesis. The fact that an infinite number of sentences are not expressed, plays no role here. The non- appearance of a particular set, which would be expected in a given situation, is the decisive criterion for the rejection of the hypothesis, not the non- appearance of some sets sometime. Cowie ( p. 223) gives an example:

Many preschoolers suspect apparently that all intransitive verbs can be used as causatives. A child hears, for example, Phrases like " I melted it" ( I melted it ) and forms analogous to the non- well-formed sentence "I giggled her" ( I chuckled ) if it wants to express that it has brought someone by tickling giggle. Suppose the child sees that his father knocked over the coffee cup so that it falls off the table. The child could now be expected based on the hypothesis that the father says: " I falled the cup off the table" ( I precipitated the cup from the table ). But this is not the case, the father says, for example, " I Caused the cup to fall from the table ." The non- occurrence of " I falled the cup off the table" in this situation is thus a negative evidence for the hypothesis of the child, that all intransitive verbs can be used as causatives.

The "repeated " POSA

According to this variant of the argument, it is not possible, from the speech input to form the rules of grammar or universal test at all. These rules are so so abstract that a prelinguistic child to any information finds in the data available to him. Thus, the universal grammar must be innate.

An already mentioned example of a part of the universal grammar is the pro -drop parameter ( see above). The knowledge that there are languages ​​in which you can not leave the subject, and languages, where that is not possible, the child can not be inferred from the linguistic input.

This variant of the POSA based on the fact that the validity of universal grammar is assumed to be proved. The example of the pro- drop parameters it can be shown, however, that the existence of the postulated by Chomsky switch is more than doubtful. So all children beginning their language development almost as if their language a null -subject language (" Will cookie! "). If, as Chomsky and other nativists say never get negative feedback for how they do it then to set the pro- drop parameter again right? In addition, in non-zero -subject languages ​​of the linguistic input contains many statements in which the subject is missing ( "must go ", "Believe not a word " or "Could not give a damn ," etc.).

Thus, if the existence of universal grammar is doubtful should not be occupied also how they can be acquired by the child.

Enlightened empiricism as an alternative to nativism

Cowie (p. 196 ) summarizes the empirical results together so that the stimulus is in fact not as "poor ," as Chomsky would have us believe. Cowie is the nativism against the enlightened empiricism. The enlightened empiricism assumes that there is quite a bit like principles and structures that restrict the choice of language learner are. However, these principles and structures are the result of prior learning experience. Nativists tend to the difficulties that a language learner to overestimate and underestimate the resources that he can draw. The child is not as nativism subordinated to the empiricism as a tabula rasa before each new step in the development of language. It uses an efficient manner to be prior knowledge in order to extract from the input meaningful rules. Instead of approaching the position of enlightened empiricism is so Cowie (p. 197), the strategy of nativists it another. Once it has been shown empirically that a particular grammatical rule can be quite learned using general learning mechanisms, is simply asserted, another rule or another principle is "not - learn ". As it is likely to persist for a long time, the debate about the POSA.