The parameter
Standard negation is the ordinary way a language negates a declarative main clause with a verbal predicate. The parameter records what kind of element does the negating: a free particle, an affix on the verb, or a verb of its own that carries the inflection the lexical verb would otherwise carry.
The classification is about form, not meaning. All four types express the same logical operation, so nothing is lost at the level of propositional content. What varies is how much of the clause must be rearranged to accommodate the negator, and whether other elements have to agree with it. That second question, negative concord, is where the parameter starts to affect translation, and it is not visible in the four-way label at all.
The labels blur at their edges. Dryer treats a negative word whose status as verb or particle cannot be determined as a value in its own right, and lists languages that vary between word and affix. These are not tidy categories recording a tidy reality; they are the least misleading cuts through a gradient.
How languages vary
- Negative particle: English not, German nicht, Dutch niet, Russian ne, Polish nie, Spanish no. A free word, usually next to the verb.
- Negative affix: Turkish -mA inside the verb, between stem and tense marking; also Japanese -nai, Swahili ha-.
- Negative verb: Finnish ei, which conjugates for person and number (en, et, ei, emme, ette, eivät) while the lexical verb appears in an uninflected connegative form. Estonian and other Uralic languages behave comparably.
- Multiple negation: French ne ... pas, where two elements together carry a single negation.
Dryer's WALS chapter 112 samples 1,157 languages: 502 use a negative particle, 395 a negative affix, 119 a double or bipartite construction, 73 a negative word of undetermined status, 47 a negative auxiliary verb and 21 vary between word and affix. Particle and affix together account for the large majority. Haspelmath's chapter 115 samples 206 languages and finds 170 in which negative indefinites co-occur with predicate negation, 11 in which they preclude it, 13 with mixed behavior and 12 using a negative existential. Co-occurrence, the pattern English forbids, is the overwhelming majority.
Consequences for translation
The formal type mostly costs a clause restructuring. Moving negation from a Turkish suffix to an English particle, or from a Finnish inflected negative verb to an English auxiliary plus not, changes which element carries the person and tense marking. It is obligatory and rarely difficult.
Concord is the part that goes wrong. In Russian, Polish, Spanish, Italian and Greek a negative indefinite requires the verb to be negated as well: nobody said nothing is not emphatic in those languages, it is the only grammatical way to say nobody said anything. English merged the two negatives into a cancellation rule centuries ago, so a doubled negative in modern standard English is marked and reads either as non-standard or as litotes. Reproducing a Slavic or Romance double negative in English therefore adds a rhetorical figure the source did not contain. The correct rendering uses one negative and an any-series indefinite, and nothing is lost, because the second source negative carried no emphasis.
The reverse direction is easy to miss. Genuine English litotes, "not unreasonable", "not without merit", is a deliberate figure, and rendering it with a plain positive erases it. That is where negation of the opposite becomes a technique rather than a repair.
Two practical notes. Scope is not preserved by word order: a negator adjacent to the verb in the source may need to attach to a quantifier or an adverbial in the target, and "all the applications were not approved" is ambiguous in English in a way the source usually is not. And a reviser working monolingually can flag a double negative but cannot tell whether the source was emphatic, which is one of the few negation problems that genuinely needs the source.
Examples
TR>EN
- Source: Başvuru süresinde yapılmamıştır.
- Target: The application was not filed in time.
- Comment: Negation is a suffix inside the Turkish verb, ordered before the tense and person marking. English carries it on a particle attached to an auxiliary. The clause is reassembled and nothing else changes.
FI>EN
- Source: En tiedä.
- Target: I do not know.
- Comment: Finnish inflects the negator for person, and the lexical verb appears in a bare connegative form. The marking English puts on the pronoun and the auxiliary sits on the negative verb.
RU>EN
- Source: Никто ничего не сказал.
- Target: Nobody said anything.
- Comment: Three negative elements, all obligatory, none emphatic. "Nobody said nothing" reproduces the form and invents a register the source did not have. One negative plus an any-form is the neutral equivalent.
References
- Dryer, M. S. (2013). "Negative Morphemes." In M. S. Dryer & M. Haspelmath (eds), The World Atlas of Language Structures Online, chapter 112. Leipzig: Max Planck Institute for Evolutionary Anthropology.
- Haspelmath, M. (2013). "Negative Indefinite Pronouns and Predicate Negation." In WALS Online, chapter 115.
- Miestamo, M. (2005). Standard Negation: The Negation of Declarative Verbal Main Clauses in a Typological Perspective. Empirical Approaches to Language Typology 31. Berlin: Mouton de Gruyter.
- Haspelmath, M. (1997). Indefinite Pronouns. Oxford: Clarendon Press.
- Payne, J. R. (1985). "Negation." In T. Shopen (ed.), Language Typology and Syntactic Description, vol. 1. Cambridge: Cambridge University Press.