Antidote’s advanced corrector carries out a complete grammatical analysis of each sentence in your text, from which it delivers a set of multifaceted diagnostics.
The scope of Antidote’s corrector is quite broad: it corrects agreement errors (subject–verb, noun–determiner, etc.), errors with the possessive ’s, commonly confused words (its/it’s, your/you’re, who’s/whose, their/there/they’re, etc.), contextual errors involving homophones or near-homophones (cup of *tee, summer *brake, world *piece). It corrects the use of hyphens, commas and capitalization, as well as the written form of numbers, times and addresses. It flags redundancies, inconsistencies, false friends, regional words, and words of inappropriate language registers. It presents its diagnostics in three distinct views: language, typography and style, for a more streamlined correction process. You can also adjust the nature and number of Antidote’s alerts by means of detailed settings.
Antidote’s corrector seamlessly integrates with most writing software. You can use it to analyze a sentence, a paragraph or two, or an entire text, depending on your needs, without changing your writing environment.
The corrector interacts closely with Antidote’s other resources. When it detects an error, it can open the appropriate guide article describing the rule that’s been broken. You can also consult the dictionary entry of a word flagged by the corrector.
Language correction requires a set of rules to follow. In establishing its rules, Antidote’s corrector aims to apply the standard conventions of English usage, as described in authoritative works such as The Cambridge Grammar of the English Language, the Oxford English Grammar, The Canadian Style, New Hart’s Rules and The Chicago Manual of Style. When these sources disagree, Antidote usually offers settings to adapt the corrections to the user’s preference.
But Antidote’s work goes well beyond grammar. It implements detailed verifications of typographic rules, from date formats to quote styles. And it goes as far as helping you improve the style of a text by highlighting repetition, problematic constructions, and vocabulary weaknesses. Moreover, using the results of the system’s language analysis, revision filters visually illustrate certain pragmatic aspects (Who, When, Where, How Many/Much), logical relations (transitions, reporting clauses), in addition to certain semantic aspects of the text (whether its connotation is positive or negative, strong or weak). Inspection filters expose each sentence’s constituent elements: adjectives, adverbs, complements, conditionals, etc. Statistics filters aggregate various aspects of language and represent them graphically: the number of errors, the most frequent words, word distribution according to etymology, etc. By submitting a text to the corrector and then to its various filters, one can examine, correct, revise and refine the text to a degree hitherto unattainable.
Despite its cutting-edge technology, however, Antidote’s corrector is not omniscient. Although each new edition brings significant progress, the expressive complexity of human language remains an enormous technical challenge. Although it can analyze a text syntactically, detecting many grammar and spelling errors, and even certain constructions that make no sense, Antidote does not understand the actual meaning of the text. It cannot rephrase a poorly written sentence, and will inevitably flag false detections and fail to catch a few mistakes.
Antidote’s corrector is a rich and valuable tool for writers, but it does not replace a human editor, and vice versa. The best way to check a text is to begin with Antidote, which flags the most embarrassing errors, including those that can easily escape our attention, and then call in a human editor who will highlight the more difficult problems. Sometimes, however, a human editor may not be available, whereas Antidote is always at the ready.