GRAPH-WIDE SEARCH
MEASURE THE GRAPH ↗NOT JUST A DICTIONARY LOOKUP
6 EVIDENCE COLLECTIONSSEARCH
THE GRAPH.
Find a modern word—or search sideways through reconstructed roots, historical forms, meanings, languages, scholars, sources, claims, disputes, and witnesses.
WIKTIONARY VIA KAIKKI / WIKTEXTRACT
Imported dictionary evidence
These are source reports, not reviewed *SET-U ancestry. A list of templates or numbered etymologies is not an ancestor chain. Scholarly disagreements remain in the original source text.
35,981 English word records · 45,491 etymology sections in this online subset
PARTIAL ENGLISH COVERAGE · FULL CORPUS REMAINS LOCAL3 source sections across 1 headword matching “Cog”. These are separate from the reviewed results below.
cogEnglish (en) · Etymology 1Source text
Etymology tree Proto-Indo-European *gēw- Proto-Indo-European *gugāder. Proto-Germanic *kuggōder. Old Norse *koggrbor. Middle English cogge English cog Inherited from Middle English cogge, from Old Norse *kogge, *koggr (see Old Swedish kogge, kogger), from Proto-Germanic *kuggō (“cog, swelling”), from Proto-Indo-European *gugā (“hump, ball”), from *gēw- (“to bend, arch”). Compare Lithuanian gugà (“pommel, hump, hill”). Cognates includes: Swedish kugg, kugge (“cog tooth”), Norwegian kugg (“cog”). The meaning of “cog” in carpentry derives from association with a tooth on a cogwheel. Compare Old Swedish koggavidher (“cog wood”), “wood reserved for a millwheel”. See also dialectal English cag (“stump”), keg; Old Norse kaggi (“keg”) + -gi (diminutive suffix), from the Germanic base *kagô (“bush, branch, stalk, stump”); also found in Bavarian Kag (“the stalk or stem of a cabbage”); dialectal Swedish kage (“treestump; piece of wood; post”), kagg or kagge (“scythe handle”); Norwegian Nynorsk kage or kagge (“low lying bush, small tree”), dialectal kagg (“scythe handle”); Old English ċeacga (“broom, furze, gorse”), whence English chag (“branch”), also Old English cyċġel, English cudgel (“knotty club”). The ultimate origin could be related to English cog (“cargo boat”) (Dutch kogge), probably named for its “round swollen” appearance.
Original Wiktionary entry and references ↗cogEnglish (en) · Etymology 2Source text
From Middle English cogge, from Middle Dutch kogge, cogghe (modern kogge), from Proto-Germanic *kuggō, from Proto-Indo-European *gugā (“hump, ball”) (compare Lithuanian gugà (“pommel, hump, hill”)), from *gēw- (“to bend, arch”). See etymology 1 above.
Original Wiktionary entry and references ↗cogEnglish (en) · Etymology 3Source text
Uncertain. Both verb and noun appear first in 1532.
Original Wiktionary entry and references ↗INDEPENDENT SOURCE LIBRARY
More sources for this spelling
These are spelling matches, not a claim of identical senses or agreement. Historical accounts retain their own editions and limits; none is automatically promoted into reviewed ancestry.
Mayhew–Skeat (1888)
1888; pinned HTML transcription updated October 28, 20240112
MODERN ENTRY POINTS
WORD DOSSIERS
Middle Irish noundossier summary
cais
eDIL separates two Irish headwords written cais. The first, also recorded as caise, is explicitly explained both as love and as hate; its cited passages and glossary equations range from searc and grádh "love" to miscais and mioscais "hatred", with context deciding individual readings. One Éigse line is tentatively translated so that the same visible form occurs on both sides: "my love ... is met by hatred(?)". The question mark matters. The second eDIL headword cais means "eye" and is excluded from this semantic lineage. These records establish a remarkable lexicographic opposition, not one timeless essence, a chronology from love to hate, or proof that every occurrence belongs to one historically indivisible sense. John T. Koch compares the noun with Proto-Celtic *katsi- < *kad-t/si- and a wider *kad- family, but that ancestry and its proposed Germanic comparison remain open. eDIL itself generally supplies etymologies only selectively. No manuscript pixels were directly inspected, no earliest-use or historical-frequency claim is made, and no exact pronunciation is invented.
OPEN THE FULL LINEAGE ↗English common noun and historical adjectivecontrarian claim
friend
English friend continues Middle English frẹ̄nd and Old English frēond, an inherited Germanic nd-stem agent noun. OED identifies cognates across West, East, and North Germanic and analyzes them as noun uses of a present participle of the Germanic base behind free; Bosworth–Toller explicitly calls Old English freónd the participle of freón, a verb meaning free or love; American Heritage independently explains the formation as originally one who loves. The historical noun later covers intimate companion, ally, supporter, relation, lover, form of address, Quaker, climbing device, and social-network contact. The graph displays a normalized *frijōnd- participial formation, Germanic *frī-, the proposed Indo-European comparison *prī-, and then stops.
OPEN THE FULL LINEAGE ↗Old Norse masculine nouncontrarian claim
frændi
Old Norse frændi belongs to the same inherited Germanic present-participial family as English friend, German Freund, and Dutch vriend, but cognacy did not freeze one meaning. The Dictionary of Old Norse Prose structures 225 citations into friend; foster brother or companion; kinsman, relative, or son; parents; and ancestor, while warning that the friend and kin divisions are not sharp. Cleasby and Vigfússon call the Scandinavian semantic specialization striking and treat kinsman as the ordinary center; Zoëga likewise gives kinsman, brother, son, and ancestor. Fritzner preserves the historical formation as a participle of frjá and compares the West and East Germanic friend words. The graph joins frændi directly to the same reviewed *frijōnd-, *frī-, and proposed *prī- nodes as friend, Freund, and vriend, then stops. The displayed pronunciation is a normalized approximation, not recorded medieval audio.
OPEN THE FULL LINEAGE ↗Old Norse weak verbdossier summary
hata
Old Norse hata is not only a medieval label for an inward feeling. The Dictionary of Old Norse Prose structures 101 prose citations under four active senses before its participial and reflexive branches: persecute; despise, reject, or cast aside; mock or humiliate; and hate, despise, or stand hostile toward. That ordering is a lexicographic analysis, not a chronology of semantic change. Fritzner independently begins with pursue or seek to destroy, continues through damage or destruction, and only then gives hate opposed to love. Zoëga likewise preserves poetic damage or destruction beside ordinary hate, while Cleasby and Vigfússon record active, reflexive, reciprocal, participial, and disputed poetic uses. Lexicon Poeticum displays eleven linked rows, including variants, damage, destruction, persecution, loathing, haters, and the gold-spending formula, but warns that its concordance has not been reviewed and that dates need checking. A c1200 scan of Holm perg 15 4°, folio 89v, was directly inspected against ONP's diplomatic transcription and visibly preserves hata in the line Elſcaþu hréinlife oc hata engi maɴ. All other rows remain source-reported. The generated possible runic inscriptions on modern dictionary transcriptions are excluded, and no exact historical pronunciation is invented.
OPEN THE FULL LINEAGE ↗Old Frisian weak verbdossier summary
hatia
Old Frisian hatia is a weak verb whose lexicographic field extends beyond an inward feeling. Hofmann and Popkema give hate or reject, damage, and a feud-language expression for harming or pursuing someone as an enemy. A May 2026 Corpus Oudfries search for word or lemma hatia returns six records in 235,462 tokens, but those are not six instances of this verb. Buma's critical note on Brokmer B 101 explicitly identifies its written hatia as atia "rear" with prothetic h, and Nijdam translates the passage as a poor mother being unable to feed her children. That token is excluded from the hate lineage. The five relevant corpus records occur in legal, road, family, and theological contexts: forms hatijen, hatie, hata, and hatien can concern harming or hindering on a road, a writing or event going against someone, or God hating perjury, lying speech, and false witnesses. The current corpus count is a bounded research surface, not a frequency estimate for medieval speech or a claim about every surviving witness. The underlying manuscripts were not directly inspected for this dossier, no earliest-use claim is made, and no exact historical pronunciation is invented.
OPEN THE FULL LINEAGE ↗Old Saxon class-II weak verbhistorical form
hatōn
The normalized Old Saxon weak-verb lemma hatōn is supported by a four-locus dictionary inventory: Heliand 1451 writes infinitive hatan in manuscript M and haton in manuscript C; Heliand 4915 writes genitive-plural present-participial hatandiero in M and hettendero in C; manuscript C supplies preterite hatoda at Heliand 5423; and GlG supplies infinitive haton aligned with Latin odisse. Those are six witness readings at four cited loci, not six independent lexemes or four manuscripts. The first Munich infinitive and the Munich hatandiero locus were directly inspected. The category problem is deeper than spelling: Köbler also indexes the Heliand 4915 forms under the separate agent noun hėttiand, enemy. The record therefore exposes the exact point where a live participial form and a lexicalized enemy noun overlap. It does not prove that pursue was the root meaning, that every enemy hates, or that every indexed form belongs unambiguously to one synchronic category.
OPEN THE FULL LINEAGE ↗Old English weak-verb lemmahistorical form
hettan
An Old English weak-verb lemma supported by a strikingly narrow surface record. A full-corpus lemmatisation inventory lists only hettende under hettan; Bosworth–Toller redirects the bare headword to hettend, “enemy,” while Orel compares the verb with Gothic hatjan and Middle High German hetzen. Clark Hall instead assigns hettende to hatian. The dossier therefore separates the comparative verb, the recorded enemy word, and the unresolved lemma assignment.
OPEN THE FULL LINEAGE ↗nouncontrarian claim
hostis
The Latin word for an outsider or public enemy, preserving the historical semantic turn from foreign stranger to armed opponent.
OPEN THE FULL LINEAGE ↗nouncontrarian claim
kayık
The Turkish word for a small boat, continued from Old Turkic kayguk and analyzed within an internal Turkic history rather than the Inuit qayaq family.
OPEN THE FULL LINEAGE ↗nouncontrarian claim
mouse
The inherited English name of the small rodent, continued from Old English mūs and the reconstructed Indo-European *mūs- family.
OPEN THE FULL LINEAGE ↗Dutch masculine de-nouncontrarian claim
vriend
Dutch vriend continues Middle Dutch vrient and Old Dutch friunt. EWN identifies Old Dutch records around 1100 and derives the word from Proto-Germanic *frijōnd-, the old present participle of *frijōn-, to love; WNT and MNW independently preserve the participial analysis and a much wider historical sense field including companion, relative, ally, patron, lover, address, mediator, and supporter. The modern spelling has one syllable and the pronunciation /vrint/; historical vrind and vrund variants prevent spelling from being mistaken for unchanged sound. The graph reuses the same reviewed *frijōnd-, *frī-, and proposed *prī- nodes as English friend and German Freund, then stops.
OPEN THE FULL LINEAGE ↗noundossier summary
word
A meaningful spoken or written unit; an inherited English form whose Germanic branch is cognate with Latin verbum rather than borrowed from it.
OPEN THE FULL LINEAGE ↗022
ATTESTED / RECONSTRUCTED / BOUNDARY
HISTORICAL FORMS
0324
WHERE INTERPRETATION IS TESTED
CLAIMS & DISPUTES
SCHOLARLY DISPUTES / 12
openproposedposition text
Do Celtic cais and Greek κῆδος prove that love and hate share one root?
No. They belong to a disputed wider comparison. Middle Irish cais carries a striking love-and-hatred field, but that does not make German Liebe an ancestor or cognate of Hass.
openprobableposition text
What securely precedes Old English brȳd?
The sources publish broad Germanic comparanda, including Old High German brūt, but this dossier does not yet have a reviewed source that warrants one reconstructed parent.
openproposedposition text
Are the opposite readings one polysemous word, merged historical homonyms, or a lexicographic convenience?
eDIL groups them in one headword but does not, by that editorial act alone, settle the full historical relationship among every love and hate occurrence.
openproposedposition text
Does cais securely prove cognacy with the Germanic hate family?
No. The Celtic–Germanic comparison is serious enough to display, but the graph keeps it as a context link rather than a genealogical edge because formal and semantic acceptance remain unsettled.
openproposedposition text
Is *kand- a secure endpoint?
No. The Latin candēre family is attested, but its deeper reconstructed root and cognate set remain analysis-dependent.
openestablishedposition text
Was emoticon first used in 1987 or only by 1992?
Merriam-Webster reports first known use in 1987; Etymonline reports the word by January 1992 and quotes an August 1992 explanation.
openestablishedposition text
Was fjá obsolete at one exact medieval date?
No. Cleasby calls the word obsolete from a nineteenth-century lexicographic perspective, but the reviewed record does not identify a last speaker or extinction year.
openproposedposition text
Was hatan inherited unchanged from Proto-Germanic?
Possibly, not securely. Orel publishes *xatēnan ~ *xatōjanan behind a Germanic cognate set. Sturtevant instead judges a specifically Gothic secondary origin more probable, and Braune records fluctuation with hatjan.
openprobableposition text
Does hatan securely descend from *ḱeh₂d-?
Only probably. The Germanic family is stronger than the deeper proposal; grief, care, mourning, hatred, pursuit, and persecution do not form one transparent semantic sequence, and some proposed cognates may not belong together.
openprobableposition text
Does hate securely descend from *ḱeh₂d-?
Only probably. The formal correspondence is widely published, but grief, care, mourning, love, and hatred do not form a transparent semantic chain, and some proposed cognates may not belong together.
openprobableposition text
Why does Köbler print hatia with an asterisk while corpus forms are visible?
Köbler's sigla mark inferred or unattested representations inside his lexicographic system, while the later corpus surfaces inflected forms hatijen, hatie, hata, and hatien and one misleading hatia homograph. The exact source-base difference remains to be reconciled.
openestablishedposition text
Can one exact Old Frisian pronunciation or vowel length be recovered?
No. Lexicographic forms include hatia, hetia, and hātia, but the reviewed sources do not provide audio or a complete dated phonetic transition series.
CONTRARIAN TRUTHS / 12
RANK 36GRADE Aranked reveal
German Feind is English fiend’s undemonic cousin: both fossilize one who hates.
Duden, DWDS/Pfeifer, and Grimm independently carry Feind through Middle and Old High German to a substantivized hating participle. OED places English fiend on the collateral Old English branch. German retained the ordinary enemy or opponent at its semantic center; English turned the cognate toward Devil, demon, and vicious person.
RANK 52GRADE Acontrarian claim
Friend and fiend are grammatical twins—but not cognates.
OED analyzes friend as a noun use of the present participle behind the Germanic free and love family, and it explicitly calls fiend a similar formation. Bosworth–Toller makes the Old English structure visible: freónd is the loving participle of freón, to free or love. The enemy word follows a different hate verb. The real unveiling is structural: one inherited agent noun is the loving one, the other the hating one or enemy, and parallel grammar must not be mistaken for one root splitting into moral opposites.
RANK 54GRADE Acontrarian claim
vriend and vijand are loving-one and hating-one mirror fossils—not cognates.
EWN, WNT, and MNW independently analyze vriend as an old present participle of the Germanic verb to love. EWN explicitly compares the formation of vijand, whose reviewed ancestry follows a different hate verb. Modern /vrint/ also makes visible how far living pronunciation can move from historical spelling. The real truth is a shared grammatical blueprint with distinct roots, not a manufactured story of one word dividing into friend and enemy.
RANK 55GRADE Acontrarian claim
frændi is friend’s cognate—but kinship became its Old Norse center.
The Germanic family did not carry one frozen dictionary definition into every daughter language. OED places Old Icelandic frændi beside English friend; Cleasby–Vigfússon and Fritzner preserve the old participial analysis; ONP then exposes what happened in use. Its 225 citations distinguish friend, foster brother or companion, kinsman or son, parents, and ancestor, with no sharp wall between friend and kin. The unveiling is semantic divergence backed by a shared graph: English centers chosen companion, while the Scandinavian cognate centers family without ceasing to be cognate.
RANK 59GRADE A-ranked reveal
Old Norse hatr has a 54-to-1 grammar problem.
The University of Copenhagen’s Dictionary of Old Norse Prose separates a 54-citation neuter noun from a one-citation masculine analysis—and flags the lone masculine assignment as possibly an elementary error. A directly inspected c.1150-1200 manuscript visibly preserves Hatr, but a visible token cannot settle gender by itself. The reveal is not that Old Norse had two equally secure hate nouns; it is that rigorous lexicography preserves the fracture instead of hiding it.
RANK 76GRADE Acontrarian claim
The perfect-looking boat cognate is false.
Greenlandic qajaq and Turkish kayık resemble one another and name similar craft, but their documented prehistories separate into Proto-Eskimo and Old Turkic.
RANK 84GRADE Aranked reveal
Even the source-form can wear an asterisk.
French lexicography defines an étymon as attested or reconstructed/hypothetical. The unit used to explain a word’s history is not automatically a surviving word.
RANK 137GRADE A-ranked reveal
Middle Irish cais could mean love or hate—and a separate cais meant eye.
eDIL's first numbered cais headword is explicitly explained as both love and hate. Its glossary evidence equates the form with serc and grádh on the love side and miscais or mioscais on the hatred side; cited passages likewise cross the divide. Most strikingly, one Éigse line is tentatively translated "my love ... is met by hatred(?)", placing the same visible form on both sides while preserving the editor's question mark. Then eDIL's second numbered cais headword breaks the spell entirely: it means eye. The unveiling is not that love and hate were secretly identical. It is that written identity, lexicographic grouping, contextual meaning, and historical ancestry are four different claims.
RANK 149GRADE Aranked reveal
English mouse is not a Latin loan.
English mouse follows its inherited Germanic line while Latin mūs is a cognate branch. Their resemblance reflects common ancestry, not a direct Latin-to-English borrowing.
RANK 163GRADE Aranked reveal
The enemy was once the stranger.
Older Latin hostis named a foreigner or legally recognized outsider. Roman authors themselves explain how usage narrowed toward the armed or public enemy.
RANK 194GRADE A-contrarian claim
A cognate can strengthen a reconstruction without proving it.
Armenian ստոյգ “true, real” is one reason the Greek *set-u- analysis is attractive; its own preform and the exact common morphology remain reconstructed.
RANK 209GRADE A−ranked reveal
Intellect belongs to the gather-and-select family.
Latin morphology connects understanding with inter- and legere, the gather, select, and read family—but the historical pieces are not a complete theory of cognition.
043
BOOKS / DICTIONARIES / ARTICLES / DATABASES
SOURCES
historical dictionary6 CLAIM LINKS
Чай — Этимологический словарь русского языка
Max Vasmer / 1953
publisherAUDIT EVERY USE ↗lexical database5 CLAIM LINKS
Ստոյգ, ուգաց — Enriched Dictionaries of Armenian
Calfa Armenian Lexicography Project Editors / 2026
author or editorAUDIT EVERY USE ↗research database3 CLAIM LINKS
IE-CoR Cognate Set 1529: λέγω
IE-CoR Editors / 2026
source titleAUDIT EVERY USE ↗0512
FIRST USES & HISTORICAL OBJECTS
ATTESTATION WITNESSES
SOURCE REPORTEDlexicographic first use
English lexicographic first-use record for mussel
mussel / mussel / before the 12th century
Mussel — Word Historyreporting source / OPEN RECORD ↗SOURCE REPORTEDlexicographic first use
English lexicographic first-use record for mouse
mouse / mouse / before the 12th century
Mouse — Word Historyreporting source / OPEN RECORD ↗SOURCE REPORTEDlexicographic first use
English lexicographic first-use report for December
December / December / before the 12th century
December — Word Historyreporting source / OPEN RECORD ↗SOURCE REPORTEDlexicographic first use
English lexicographic first-use record for guest
guest / guest / 13th century
Guest — Word Historyreporting source / OPEN RECORD ↗SOURCE REPORTEDlexicographic first use
Middle English lexicographic first-use record for algorism
algorisme / augrim / algorithm / 14th century
Algorism — Word Historyreporting source / OPEN RECORD ↗SOURCE REPORTEDlexicographic first use
English lexicographic first-use record for muscle
muscle / muscle / 14th century
Muscle — Word Historyreporting source / OPEN RECORD ↗SOURCE REPORTEDlexicographic first use
English lexicographic first-use record for host in the hospitality sense
host / host / 14th century
Host — Word Historyreporting source / OPEN RECORD ↗SOURCE REPORTEDlexicographic first use
English lexicographic first-use record for hospital
hospital / hospital / 14th century
Hospital — Word Historyreporting source / OPEN RECORD ↗SOURCE REPORTEDlexicographic first use
English lexicographic first-use report for nice
nice / nice / 14th century
Nice — Word Historyreporting source / OPEN RECORD ↗SOURCE REPORTEDlexicographic first use
English lexicographic first-use report for science
science / cience / science / 14th century
Science — Word Historyreporting source / OPEN RECORD ↗SOURCE REPORTEDlexicographic first use
English lexicographic first-use report for check
check / check / 14th century
Check — Word Historyreporting source / OPEN RECORD ↗SOURCE REPORTEDlexicographic first use
English lexicographic first-use report for chess
chess / chess / 14th century
Chess — Word Historyreporting source / OPEN RECORD ↗06
SEARCH IS DISCOVERY, NOT PROOF
EVERY RESULT OPENS A PERMANENT EVIDENCE RECORD.
Matching text only locates a record. It does not raise confidence, settle a dispute, establish cognacy, or turn a source report into direct inspection. Follow the result into its citations and editorial limits.