 | This articleneeds additional citations forverification. Please helpimprove this article byadding citations to reliable sources. Unsourced material may be challenged and removed. Find sources: "Cognate" – news ·newspapers ·books ·scholar ·JSTOR(December 2023) (Learn how and when to remove this message) |
Inhistorical linguistics,cognates orlexical cognates are sets ofwords that have been inherited in direct descent from anetymological ancestor in a commonparent language.[1]
Becauselanguage change can have radical effects on both the sound and the meaning of a word, cognates may not be obvious, and it often takes rigorous study of historical sources and the application of thecomparative method to establish whetherlexemes are cognate. It can also happen that words which appear similar, or identical, in different languages, arenot cognate.
Cognates are distinguished fromloanwords, where a word has been borrowed from another language.
The English termcognate derives fromLatincognatus, meaning "blood relative".[2]
For an example, cognates with the English wordnight can be found in most majorIndo-European languages, includingGermanNacht,Swedishnatt,Ukrainianнічnich,Russianночьnoch,Lithuaniannaktis,Welshnos,Polishnoc,Greekνύχταnychta,Sanskritनक्तnakt-,Albaniannatë,Latinnox (gen. sg.noctis),Italiannottecode: ita promoted to code: it,Frenchnuit, andPortuguesenoite. These all mean 'night', and derive from the Proto-Indo-European*nókʷts with the same meaning. The Indo-European languages have hundreds of such cognate sets, though few of them are as neat as this.
TheArabicسلامsalām, theHebrewשלוםshalom, theAssyrian Neo-Aramaicܫܠܡܐshlama andAmharicሰላምselam 'peace' are cognates, derived from theProto-Semitic*šalām- 'peace'.
TheParaguayan Guaranipanambi, theEastern Bolivian Guaranipanapana, theCocama andOmaguapanama, and theSirionóana ana are cognates, derived from theOld Tupipanapana 'butterfly', maintaining their original meaning in theseTupi languages.Brazilian Portuguesepanapanã (flock of butterflies in flight) is a borrowing rather than a cognate of the other words.
Cognates need not have the same meaning, as they may have undergonesemantic change as the languages developed independently. For exampleEnglishstarve andDutchsterven 'to die' orGermansterben 'to die' all descend from the sameProto-Germanic verb,*sterbaną 'to die'.
Cognates also do not need to look or sound similar: Englishfather,Frenchpère, andArmenianհայր (hayr) all descend directly fromProto-Indo-European*ph₂tḗr. An extreme case is Armenianերկու (erku) and Englishtwo, which descend fromProto-Indo-Europeandwóh₁; the sound change*dw >erk in Armenian is regular.
Paradigms of conjugations or declensions, the correspondence of which cannot be generally due to chance, have often been used in cognacy assessment.[3] However, beyond paradigms, morphosyntax is often excluded in the assessment of cognacy between words, mainly because structures are usually seen as more subject to borrowing. Still, very complex, non-trivial morphosyntactic structures can rarely take precedence over phonetic shapes to indicate cognates. For instance,Tangut, the language of theXixia Empire, and oneHorpa language spoken today inSichuan, Geshiza, both display a verbal alternation indicating tense, obeying the same morphosyntactic collocational restrictions. Even without regular phonetic correspondences between the stems of the two languages, the cognatic structures indicate secondary cognacy for the stems.[4]
False cognates are pairs of words that appear to have a common origin, but which in fact do not. For example, Latinhabēre and Germanhaben both mean 'to have' and are phonetically similar. However, the words evolved from differentProto-Indo-European (PIE) roots:haben, like Englishhave, comes from PIE*kh₂pyé- 'to grasp', and has the Latin cognatecapere 'to seize, grasp, capture'.Habēre, on the other hand, is from PIE*gʰabʰ 'to give, to receive', and hence cognate with Englishgive and Germangeben.[5]
Likewise, Englishmuch and Spanishmucho look similar and have a similar meaning, but are not cognates:much is from Proto-Germanic*mikilaz < PIE*meǵ- andmucho is from Latinmultum < PIE*mel-. A true cognate ofmuch is the archaic Spanishmaño 'big'.[6]
Cognates are distinguished from other kinds of relationships.
Anetymon, or ancestor word, is the ultimate source word from which one or more cognates derive.In other words, it is the source of related words in different languages. For example, the etymon of both Welshceffyl and Irishcapall is the Proto-Celtic*kaballos (all meaninghorse).
Descendants are words inherited across a language barrier, coming from a particular etymon in an ancestor language. For example, Russianмо́ре and Polishmorze are both descendants of Proto-Slavic*moře ('sea').
Aroot is the source of related words within a single language (no language barrier is crossed).
Similar to the distinction betweenetymon androot, a nuanced distinction can sometimes be made between adescendant and aderivative.
Aderivative is one of the words which have their source in a root word, and were at some time created from the root word using morphological constructs such as suffixes, prefixes, and slight changes to the vowels or to the consonants of the root word. For exampleunhappy,happily, andunhappily are all derivatives of the root wordhappy.
The termsroot andderivative are used in the analysis of morphological derivation within a language in studies that are not concerned with historical linguistics and that do not cross the language barrier.
{{cite web}}: CS1 maint: location (link)