Where’s the sense in double negatives?

The use of double negation in English persists despite being counterintuitive, suggests a study by Mora Maldonado and Jennifer Culbertson at the University of Edinburgh in Scotland.

English children’s misuse of double negatives is constantly corrected, such as “I never said nothing!”, intended to mean “I never said anything!”, but in fact meaning the opposite, as English follows the rule of double negation, ie, two negatives make a positive. Many languages, however, do not follow this rule but exhibit negative concord, ie, two negatives have an overall negative meaning. For example, in Serbian, “Niko ne trči” could be literally translated as “no one [not] runs” – a double negative in English, but the meaning in Serbian is “nobody runs”.

It has been difficult to establish which pattern is more natural. Germanic and Romance languages, for example, have changed back and forth between double negation and negative concord over time.

One theory proposes that the meaning depends on the kind of negative marker, so that phonologically strong markers, such as “niet” in Dutch show double negation, while the weaker Serbian “ne” shows negative concord. But this simple model has become increasingly complicated by the role of particles, adverbs and other items.

To try to settle the question of what determines how negatives are interpreted, Maldonado and Culbertson developed four artificial languages that varied in two ways. First, they varied by type of negative marker, being either an affix (an addition to the stem word, such as the prefix ‘dis’ or suffix ‘less’ in English) or an adverb (such as the English ‘not’ or ‘never’).

Second, languages varied by having a double negative or negative concord interpretation. Overall, this gives the four language types. These are: (1) affix or (2) adverb marker leading to double negation; (3) affix or (4) adverb marker leading to negation concord.

One hundred and twenty-four English speakers, divided into four groups, took part in the experiment, first learning simple affirmative sentences, then being tested on production and comprehension. In the second session, participants also learned simple negative sentences and again were tested. In a third session, the use of quantifiers for ‘all’ or ‘none’ were added and tested. Then, in a final fourth session, both simple sentences and sentences using quantifiers were presented. In this last session, the acquired language items made possible the target sentence forms containing two negatives which participants were asked to translate into English.

According to the theory above, English speakers should learn a language using double negation more easily when the negative marker is an adverb and conversely, using an affix for negation should lead naturally to negative concord as a general rule. If this were true, it might shed further light on language

learning models by implying a cognitive constraint during language learning.

Analysis of the scores from session four, however, found that whether the negative marker was an affix or an adverb did not influence accuracy scores, but scores were significantly higher when learning the languages using negative concord rather than double negation. Learners were also slower to comprehend double negation. It seems that languages using double negation are harder to learn no matter what kind of negative marker is used.

Researchers studying natural language acquisition have proposed that negative concord is the default for young learners and this study suggests this may extend into adulthood. Interpreting double negation as a positive requires assessing each negative element independently and it seems that learners tend instead to give a negative meaning to the overall sentence.

Certainly, negative sentences are generally more cognitive work to process and that cost might be less for negative concord. It may be that negative concord is actually a more natural interpretation of English – as reflected in the common ‘misuse’ of double negatives by both children and adults.

It is difficult to draw a definitive conclusion regarding the influence of affixes vs adverbs due to the simple nature of the languages used and the possible biases of the English-speaking participants, but the overall finding that, like children, adults prefer negative concord to double negation begs the question as to why languages like English persist in the latter.

Perhaps this has more to do with sociology than linguistics, as the misuse of double negatives is one of the characteristics of ‘substandard English’ that’s used to differentiate England’s notorious class system, as epitomised by Eliza Doolittle’s “I ain’t done nothing…”.

REFERENCE

• Maldonado, M and Culbertson, J (2021), Nobody Doesn’t Like Negative Concord, Journal of Psycholinguistic Research 50: 1401-1416 https://doi.org/10.1007/ s10936-021-09816-w