These lessons introduce you to the etymology of words. They build on historical linguistics.

  1. Sound & Shape Changes
  2. Meaning Changes

The Path of Sound Change & Form Change

An Overview of Etymology

Etymology explores the history and development of individual words, the origins of a language's lexical items. It asks a question you're likely familiar with: "where did this word come from?". You've probably encountered etymologies when someone explained the meaning of a word, or when you sought out the origin of a word. As a linguistic study, this question is approached methodically, so it requires that we understand the methods that we'll use to discover word origins.

Specifically, etymology requires an understanding of the parts of words - their pronunciation and grammar components. So you'll need some understanding of phonology & morphology. Visit those earlier lessons for an introduction to the grammar of words, and an intro to pronunciation in language.

Etymology is also a diachronic process. In other words, etymology is a function of a language's change over time. It will make more sense if you're already versed in the fundamentals of historical linguistics. This includes awareness of key concepts like cognates and borrowings, reconstruction of proto-languages, and discussions about the difference between lects (languages, dialects, idiolects).

With that knowledge, we can start to build etymologies. For the etymology of any particular word, we'll need to know about language change over time in general and the history of change in the particular language we're analyzing. We will also need to know the history of borrowing ("loanwords" or "loans") into that particular language (and, in turn, the historical changes that those source langs underwent).

Keep in mind that words are often borrowed or inherited from nonstandard forms, which adds to the complexity. For instance, Middle English speakers borrowed the word 'chief' from Medieval Norman French, not the Modern French 'chef'. You may also run into examples of dialect mixing, which accounts for the English doublet 'skirt' - 'shirt', both originating from the same root via different dialects.

A Sample Etymology

First, let's discuss how you can trace the path of a word as its sound and form change. As much as possible, you rely on historical attestation. Here you expect your etymology to account for the evidence, that is, the word you're tracing as it was actually used throughout its history. For example, the Modern English word 'queen' goes back to an earlier 'quene' in Middle English. This word is attested in Chaucer (CT - University of Chicago Ms. 564). 'Quene' in turn derives from Old English 'cwēn', attested in Beowulf (Nowell Codex, ca. 1000). Here we can say that the etymon (plural: etyma), or source word, of 'queen' is the Old English 'cwēn'.

language etymon attestation
Modern English queen (modern speakers)
Middle English quene (Chaucer, etc.)
Old English cwēn (Beowulf, etc.)

You aren't restricted by the attested evidence. There are two ways to trace a more distant origin. First, you might discover that a word was influenced by or borrowed from another dialect or language, like the English word 'chief', borrowed from French. In this case, the etymology can continue in that source language. Second, you can increase the time depth by comparing related languages and engaging in the process of historical reconstructions.

Therefore, as much as is possible, etymology accounts for the history of borrowing & influence, and terminates at the earliest reconstructible form of the word, not the earliest attested form. For example, the Old English word cween can be compared to other early Germanic words like Gothic 'qēns' & Old Norse 'kvæn'. Their shared, reconstructed proto-Germanic ancestral form is *kwēniz.

What's more, when comparing the Proto Germanic word to cognates in other non-Germanic Indo-European languages, like Ancient Greek γυνή (gynē) & Sanskrit ग्ना (gnā), we can begin to reconstruct a Proto-Indo-European root like *gwēn-. (In this particular case, we can see that Grimm's Law changed voiced stops to voiceless stops and voceless stops to fricatives between Indo-European and Germanic, giving us a root *kwēn beginning with voiceless /k/ rather than voiced /g/.)

Let's review what we've done so far:

  • We found the earliest attested form of 'queen' in English, compared it to words in related Germanic languages to trace the word to a reconstructed proto-word, and compared Proto-Germanic to related Indo-European languages to trace the word back to a reconstructed Proto-Indo-European root.
  • Along the way, we paid attention to sound changes, including the formidable Grimm's Law.
  • We followed common etymological conventions, like placing an asterisk before reconstructed forms and representing the derivational path visually (*kwēniz > cwēn > quene).
direction path
past to present *kwēniz > cwēn > quene
1 becomes 2 becomes 3
present to past quene < cwēn < *kwēniz
3 comes from 2 comes from 1

It's also crucial to understand morphology, to know what kind of word you're dealing with. Especially key here is "word formation" - the way affixes are added to roots to form new words (derivation), or how words are compounded to form new words (compounding). Inflectional morphemes are considered, but they're really part of the "grammar" or structure of a language, while etymology often focuses on "content" words and morphemes.

Consider one example of the role of morphology in etymology. When we traced the history of 'queen', we arrived at a Proto Gmc word *kwēniz, which has an inflectional suffix -iz attached to the root noun. In a separate instance, you might be doing the etymology of, say, the word 'illogical', which has a root log- and three derivational affixes: the suffix -ic, the suffix -al, and the prefix in- (which assimilated to il-). You can only do justice to the etymology of 'il-log-ic-al' if you understand the various morphemes, their individual etymologies and their place in the overall etymology of the word.

The next section will explore the role of semantics, particularly meaning change, in etymology, and I'll also touch on basic conventions followed when presenting an etymology.

Semantics, or Whence Cometh Meaning?

Meaning Change

So far we've followed the trail of various forms of a word as it changed over time. But just stating that, for example, 'cwēn' became 'queen' ignores one crucial aspect, an aspect that initially attracts most amateur enthusiasts of etymology: semantics. When it comes to etymology, semantics deals with the meaning of words and parts of words.

Semantics tends to be fuzzy, taking the meaning of "fuzzy" in two ways. First, the relationship between the meaning of a word and its form is generally opaque. There's nothing inherent in the sounds of the word 'queen' that make it any more queen-like than the Latin word with the same meaning - 'regina'. Second, the semantic relationship between words whose meaning is known, and even between word roots and words derived from those same roots through word formation, is approximate and variable. Consider the relationship between the meaning of 'house' and the meaning of 'home', which have very similar but not identical semantic content. The etymological 're-' in 'remain' versus 'reappear' has two different values. In these and other cases, you cannot simply set semantic relatedness to true or false - you have to be aware of and capture the gradations of meaning.

Like sound and form, meaning changes over time. We can classify meaning changes into types: abstract to or from concrete, general to or from specific, temporal to or from locative, and so on. We can also consider the origin of a meaning change or the influence of other languages. For example, calques form when speakers borrow patterns found in another language to express a meaning. Rather than borrow the Latin expression 'Via lactea' for our galaxy, English speakers calqued 'Via Lactea' as a new native term 'Milky Way'. Other times, semantic loans occur when a language uses a preexisting word or phrase to cover a borrowed meaning.

We can also talk about meaning and use. Words in use may be described as literary, poetic, archaic (sound old), obsolete (no longer used), obsolescent (going out of use), jargon (specialized use), colloquial (used in conversation), slang, preferred (used more than a similar word), and so on.

However we classify it, semantic change is not as systematic as sound change. Under the regularity hypothesis, we expect that, if the affricate *dz- > z- in a language, words that once began with dz- will now start with z-. Don't expect the same from, say, an abstract > concrete semantic change - not all abstract meanings will become concrete.

A last note about meaning change - the basic core vocabulary of a language contains words that tend not to be replaced and whose meanings do not shift as much. As the name implies, these words include basic terms like family members, terms for earth, sun, and other words learned early and used often.

Let's go back and look at our etymology of 'queen'. You saw that the word came from Old English 'cwēn'. Let's add our semantics. In Old English, that word meant "woman". What's more, the Proto-Germanic word *kwēniz and most cognates in other Indo-European languages (like Greek gynē) mean "woman" or "wife". It's safe to say that the Proto-Indo-European root *gwen- also meant "woman"! This is an example of meaning change - the meaning shifts from general "(dignified) woman" to a more specific "female monarch".

Conventions for Etymologies

As you read etymologies or build an etymology, there are some conventions worth noting:

  • The source language, word and meaning are indicated for earlier attested forms: Old English cwēn "woman".
  • Parts of speech are sometimes given, such as noun, adjective or verb: Old English cwēn (n.) "woman".
  • Language names are abbreviated, and dates and sources support more thorough etymologies: O.E. cwēn (n.) "woman" (Beowulf, 11c.).
  • The complex relationships between cognates and dialects are boiled down to clear examples with relevant semantic information given for each example, either by choosing the most representative attested word or by citing a reconstructed ancestral form: queen (n.), from O.E. cwēn "woman", from ProtoGmc *kwēniz.
  • Comparative examples are also cited: O.E. cwēn (n.) "woman", cf. O.H.G. quëna, Goth. qens.
  • Again, reconstructions are marked with an asterisk: queen, from P.Gmc. *kwēniz, from PIE *gwen-.
  • Words are traced back as far as possible. Sometimes an etymology becomes unclear or contested, in which case the uncertainty is mentioned and the multiple possible origins are cited. For instance, there are double etymologies and cases of influence: English fuse (n.), from It. fuso "spindle", influenced by Fr. fusée "hemp fiber spindle".

Where to Read & Hear Etymologies

Let's conclude this discussion of etymology by considering the contexts in which we may stumble upon etymologies.

You'll often find etymolgies in major dictionaries following definitions of many words. You can also get your hands on dictionaries of proto-language root words if you're interested in doing etymology in reverse - determining the words that have descended from some root word in a specific reconstructed language, like Proto-Indo-European or Proto-Mayan.

A common source of etymologies is folk etymology - the popular telling and retelling of the stories behind a word. Folk etymologies often focus on names and on popular phrases like 'mind your p's and q's'. Of course, folk etymologies are not always factually correct, but more importantly, they're never accurate enough to count as thorough linguistic etymologies. Folk etymologists are interested in storytelling, entertainment and credibility, not precision and evidential support.

Speaking of names, onomastics studies the etymology of proper names. The meaning of words that turn into names tend to become vestigial - the name proper comes to refer to the identity of a person or place, but the etymology still interests even non-specialists. What's more, words derived from same source can change in different ways when they're used as a proper noun. Names may be more conservative (change less) over time or capture the history of borrowings and interactions with surrounding languages. For example, the French name 'Jacques' preserves an old nominative singular case ending -s, which common nouns lost. The Spanish name Rodríguez contains a Visigothic (Germanic) root *hrod-ric with a patronymic suffix *-ez.

That's all for this introduction to etymology. I hope it's helped you begin to understand etymology from a linguistic perspective. You see, while an etymology won't tell you what word actually means (that's determined by how speakers of a language use that word), it does tell the fascinating - and sometimes unexpected - history of the development of words in a language. The next time you hear or look up an etymology, you can take a more critical and rigorous look at the origin of a word.