Medical Terminology: How Medical Words Are Built

Definition

Medical terminology is a constructed vocabulary built mainly from Greek and Latin word parts, roots, prefixes, suffixes, and connecting vowels, combined under a small set of rules, so a word's meaning can usually be worked out from its pieces instead of memorized whole. Greek roots dominate the language of disease and diagnosis. Latin roots dominate the language of anatomy. The split has more to do with history than logic.

Take the word osteomyelitis apart before you know what it means, and you already know most of what it means. Oste- is bone. -Itis is inflammation. Myel- is marrow, sitting between them like a hinge. Put it together and you get inflammation of the bone and its marrow, and you got there without a dictionary, using nothing but the pieces. That's the trick behind medical terminology roots, prefixes, and suffixes: this isn't a vocabulary to learn word by word. It's a small, reusable kit of Greek and Latin parts, assembled according to fairly consistent rules, that has been growing for roughly 2,500 years and is still being extended now.

Two languages, one accident of history

Ask why anatomy runs on Latin and disease runs on Greek, and the honest answer is that two groups of people, working centuries apart, got to name things first, and everyone after them built on what was already there.

The vocabulary of disease, what an illness does to the body, how it progresses, what to call a swelling or a wasting, goes back to the Hippocratic physicians of the fifth century BCE, working on the Greek island of Kos and around the Aegean. Britannica's entry on Hippocrates describes a tradition that organized illness through direct observation rather than through appeals to gods or omens, and the vocabulary that tradition needed is where -itis (inflammation), -osis (a condition or process), and pathos itself (suffering, disease) come from. Later Greek physicians extended that vocabulary a great deal, none more than Galen in the second century CE, who wrote in Greek even while practicing in Rome. His authority kept Greek as the working language of pathology for well over a thousand years after him, so when European medicine began coining new disease names in the Renaissance and after, it kept reaching for Greek roots because that was already the house style.

Anatomy took another path. The body's structures, bones, muscles, vessels, organs, got their lasting names not from Greek physicians but from the Latin-writing anatomists of Renaissance Europe, most famously Andreas Vesalius, whose 1543 De Humani Corporis Fabrica set out to describe the body from direct dissection rather than from what earlier authorities had written about it. Latin was the working language of European universities at the time, so Latin is what the naming happened in, even where the underlying word was originally Greek: cranium, for instance, is a Latinized form of the Greek kranion. That naming project eventually got formalized. The Basel Nomina Anatomica of 1895 tried to standardize anatomical names across countries, and the current international standard, Terminologia Anatomica, published in 1998, is still written in Latin, not because Latin has any special claim on describing bones, but because that's the language the naming tradition happened to be running in when it got codified.

The result is a vocabulary with two grammars sitting inside one profession. A bone is named in Latin: femur, tibia, clavicula. What happens to that bone gets described in Greek: osteomyelitis, arthritis, osteoporosis. Nobody designed it that way. It's what was left after two separate naming traditions ran for a few thousand years and then got stitched together into a single working language.

The build kit: roots, prefixes, suffixes, and the connecting vowel

Every constructed medical word draws on up to four kinds of pieces, and once a reader can spot which is which, unfamiliar terms stop being unfamiliar.

The root carries the core meaning: the body part, substance, or process the word is about. Cardi- is heart. Hepat- is liver. Nephr- is kidney. Derm- is skin. Neur- is nerve. A word can carry more than one root: nephr/o/lith/iasis (kidney + stone + condition of) means the condition of having kidney stones.

Prefixes attach to the front and usually narrow where something is or how much of it there is. Hypo- means below or deficient; hyper- means above or excessive, so hypothyroidism and hyperthyroidism are opposites built from the same root with opposite prefixes. Brady- (slow) and tachy- (fast) work the same way against -cardia: bradycardia, tachycardia. Peri- (around), sub- (under), and supra- (above) all do the same job of pinning down location.

Suffixes attach to the back and usually name what kind of thing the word is describing: a condition, a procedure, a field of study. -Itis is inflammation. -Osis is a condition, often degenerative rather than inflammatory. -Ectomy is surgical removal. -Otomy is cutting into. -Ostomy is creating a permanent opening. -Ology is the study of. -Algia and -dynia both mean pain, inherited from two separate Greek words that ended up doing the same job.

Then there's the connecting vowel, usually o, occasionally i, whose only function is to make two consonant-heavy pieces pronounceable when joined. Cardi + o + logy becomes cardiology, not cardilogy, because -logy starts with a consonant and needs the vowel as a bridge. Drop the connecting vowel when the next piece already opens with one: cardi + itis becomes carditis, not cardioitis, because -itis already supplies its own opening vowel. That single rule, keep the vowel before a consonant, drop it before a vowel, explains most of the spelling choices that look arbitrary to a reader meeting a word for the first time.

Word order generally runs prefix, then root or roots, then suffix, read innermost out. Gastr/o/enter/o/log/ist: stomach, intestine, study of, the person who does it, a specialist in stomach and intestinal disease. Once the slots and the vowel rule are clear, a wall of syllables turns into an assembly instruction.

Taking a word apart, step by step

Two examples, worked through slowly, show how the method holds up on words long enough to look intimidating.

Electroencephalography breaks into electro- (electricity, from Greek elektron), en- (in), kephal- (head, from Greek kephale), and -graphy (the process of recording). Read outward from the last piece: a recording process, of the head, involving electricity, which is exactly what an EEG measures. The word looks unmanageable until it's split at its joints, at which point it's four ordinary pieces stacked in a predictable order.

Hepatosplenomegaly does the same with three roots instead of two: hepat- (liver), splen- (spleen), and -megaly (enlargement, from the Greek megas, large). Read outward again: enlargement, of the spleen, and the liver, an enlarged liver and spleen described in one word instead of a sentence's worth of clauses. A chart note that had to spell out "the liver and the spleen are both enlarged" every time would take far longer to write and read; the compound word is doing the job English would otherwise need six words for.

The method generalizes past these two examples. Find the suffix first, since it usually names the category of thing the word is describing. Work backward through the roots, right to left, inserting "of" between them. Then check the front for a prefix narrowing location or degree. It won't work on every word, eponyms and borrowed terms don't follow these rules at all, but it works on enough of the vocabulary that learning the roots pays off before any individual term does.

Where the rules bend

The system is consistent enough to be useful, not consistent enough to be reliable without exceptions, and the exceptions cluster in a few predictable places.

Plurals are the first trap. Latin and Greek nouns keep their original plural patterns inside English medical vocabulary long after most English speakers stopped learning Latin declensions in school. A vertebra becomes vertebrae, not vertebras. An alveolus becomes alveoli. A diagnosis becomes diagnoses, an ovum becomes ova, a ganglion becomes ganglia, an appendix becomes appendices. Mismatch the singular and plural, "the patient has three vertebras," and it reads as an error to anyone trained in the system, even though the sentence is perfectly understandable. Some of these plurals are drifting toward regular English endings anyway: carcinomas is now more common in clinical writing than the classical carcinomata, and stomas has mostly replaced stomata outside strict anatomical Latin. The system tolerates its own erosion; it just erodes slowly and unevenly.

The second bend is purity. Nineteenth-century grammarians worried at length about "hybrid" words, compounds mixing a Latin root with a Greek suffix or the reverse, on the theory that a well-built word should come from one language throughout. That standard was never actually followed in practice. Appendicitis pairs a Latin root (appendix) with a Greek suffix (-itis). Tonsillitis does the same with tonsilla. The hybrid form isn't a corruption of some earlier pure system, historically frowned upon by grammarians but common throughout the vocabulary anyway; mixed compounding has been happening since roughly the moment Latin and Greek medical writing came into contact with each other.

The third bend is in words that were never built at all. Diabetes, cancer, coma, and fever look like they should break down into pieces the way osteomyelitis does, but they're whole borrowed terms, single units carried over from Greek or Latin without being assembled from separable parts inside English. You can trace their history, but decomposing diabetes into a root plus a suffix won't hand you a meaning the way it does for a genuine compound. The system explains most of the vocabulary. It doesn't explain all of it, and telling the compounds apart from the old borrowings is a skill learned mostly through exposure.

Tracing one root across a family of words

Take cardi-, the Greek root for heart, and watch it move through a set of words to see how the system reuses one part to build several concepts instead of a fresh word for each one.

Cardiology is the study of the heart: cardi- plus -o- plus -logy. A cardiologist is the person who practices it, same root plus -logist, one who studies. Cardiomyopathy adds a second root, myo- (muscle), to describe disease of the heart muscle specifically, as opposed to disease of its vessels or valves. Tachycardia and bradycardia swap in the prefixes for fast and slow to describe a rate rather than a structural problem. Pericardium and myocardium put the same root in the second half instead of the first, naming the membrane around the heart and the muscle of the heart, a reminder that roots don't always sit in the same slot. And cardiac, the everyday adjective, doesn't come through the usual compound-building process at all. It's a direct borrowing of the Greek adjective kardiakos, absorbed whole into English instead of a fresh compound.

That last example is worth sitting with, because it shows how the system actually grew. Nobody engineered it top-down, deciding "heart" would always take the form cardi- in every position. It accumulated: some words compound-built centuries after Greek and Latin had stopped changing as living languages, others carried over whole from ancient usage, still others coined recently for a gap nobody had needed to name before. Hepat- (liver) shows the same pattern: hepatitis, hepatology, and hepatomegaly are built compounds, while hepatic, the common adjective, is again a direct borrowing instead of a fresh compound. Once that's visible, the vocabulary stops looking like a single tidy machine and starts looking more like a shoreline shaped by different tides at different times: the underlying rock is consistent, the shape it left behind isn't uniform.

Words built somewhere else: eponyms, acronyms, and borrowings

Not every term came out of the root-and-suffix machine, and it helps to know the other three main ways a word gets in.

Eponyms name a condition, structure, or procedure after a person, usually the physician credited with describing it first, occasionally a patient. Parkinson's disease is named for James Parkinson, an English surgeon who described it in 1817. Down syndrome, Alzheimer's disease, and the Eustachian tube work the same way. You can't work out what Parkinson's disease is from the word the way you can work out what osteomyelitis is; a person's surname carries no etymological information about the condition, so you either already know it or you look it up.

Acronyms and initialisms make up a second category, and a few have drifted far enough from their origins to function as ordinary words rather than abbreviations anyone still expands mentally. AIDS (acquired immunodeficiency syndrome) is spoken as a word, not sounded out letter by letter the way HIV still is. LASIK (laser-assisted in situ keratomileusis) works the same way; almost nobody hearing "LASIK surgery" is decoding the acronym in real time. MRI and CPR, by contrast, never made that jump; they stayed initialisms, spoken letter by letter. There's no firm rule for which acronyms cross over. It seems to depend mostly on frequency: how often the word gets said out loud by ordinary speakers, doctors included.

The third category is direct borrowing from a modern language, usually because a disease or practice was first described, or first became a public concern, outside the Greek-Latin tradition entirely. Influenza comes from Italian, originally referring to the belief that outbreaks were caused by astral "influence." Malaria is also Italian, from mal aria, bad air, reflecting the old theory that the disease came from foul swamp vapor rather than a mosquito-borne parasite. Quarantine comes from the Venetian term for the forty-day isolation period imposed on ships during plague outbreaks. Kwashiorkor, a term for severe protein malnutrition, was introduced into medical English in 1935 by the physician Cicely Williams, who borrowed it from the Ga language of coastal Ghana rather than coining a Greek compound for it. None of these words were built from parts meant to describe the condition directly. They carry an old theory or an old practice fossilized inside them, still in daily use long after the theory that produced them was abandoned.

Contemporary context: retiring names, coining new ones

The vocabulary isn't a museum piece. It keeps growing, and one visible sign of that right now is the retirement of eponyms tied to physicians whose research records don't hold up.

Two cases come up often. Reiter syndrome, a form of reactive arthritis, was named for Hans Reiter, a German physician who described it in 1916 and later held a senior public health post under the Nazi government, connected to human experimentation on concentration camp prisoners. Wegener granulomatosis, a rare inflammatory blood vessel disease, was named for Friedrich Wegener, whose Nazi Party membership and wartime record came under scrutiny decades after his death. Professional medical societies moved to drop both names in favor of descriptive terms: reactive arthritis for the first, granulomatosis with polyangiitis for the second, formally adopted by the American College of Chest Physicians in 2011. Neither change altered how the conditions are diagnosed or treated. It changed who gets remembered every time a clinician writes the word.

That's one visible edge of a broader drift away from eponyms in general. Several medical journals and style guides now discourage coining new ones, favoring descriptive terms built the old compositional way, a structure a future reader can take apart instead of a name that has to already be known. The argument isn't really about any single physician's biography. It's that a word built from named parts survives being learned cold by a stranger, and a word built from a surname doesn't. The move says something about what the profession values: legibility across borders and generations over the memory of any one person.

The root-and-suffix system, meanwhile, keeps absorbing new material the way it always has, new imaging techniques and drug classes named with the same Greek and Latin toolkit that named osteomyelitis. Not everything new goes that route, though. "Long COVID" entered clinical use around 2020 as a plain-English descriptive phrase, coined largely by patients before it was adopted by clinicians and journals, a reminder that the vocabulary is still open to terms arriving from outside the classical system entirely, not only from inside it.

Spelling, geography, and the difference between the skeleton and the accent

The Greek-Latin naming system is international in a way few professional vocabularies manage: a doctor trained in Zagreb, Buenos Aires, or Manila works from the same Terminologia Anatomica and largely the same disease roots as one trained in Boston. But the spelling of that shared vocabulary splits along the same line British and American English splits on almost everything else.

British usage tends to keep the Latin transliteration of the original Greek diphthongs: anaesthesia, oesophagus, paediatric, orthopaedic, haemoglobin. American usage simplified those spellings in the nineteenth century, largely following Noah Webster's broader push to make American English more phonetic and less tied to its classical ancestry: anesthesia, esophagus, pediatric, orthopedic, hemoglobin. Neither version is more correct. They're the same word, carrying the same root, spelled according to two editorial traditions that split roughly two centuries ago and never rejoined. A reader moving between a British and an American text mostly needs to recognize oe or ae as e and move on.

What doesn't travel as cleanly is the everyday clinical vocabulary sitting on top of that shared skeleton, the shorthand, the local abbreviations, the phrases clinicians actually say to each other on a ward. The Latin and Greek naming system is standardized by international bodies precisely so the formal anatomy and pathology mean the same thing everywhere. The words people reach for at the bedside are not standardized that way; they vary from one hospital to the next and from one region to the next, even when the formal term underneath is identical. Medical vocabulary has two layers, and only one of them was built to travel. The Latin skeleton of medical terminology is remarkably stable across countries. What gets said out loud, ward to ward, is not.

Key terms

• Root, the piece carrying the core meaning: the body part, substance, or process a term is about (cardi-, hepat-, nephr-).
• Prefix, a piece attached to the front of a root, usually narrowing location or degree (hypo-, brady-, peri-).
• Suffix, a piece attached to the end of a root, usually naming a condition, procedure, or field of study (-itis, -ectomy, -ology).
• Connecting vowel, the linking vowel, usually o, inserted between a root and a suffix beginning with a consonant, and dropped when the next piece already opens with one.
• Eponym, a term named for a person, place, or myth instead of built from descriptive roots (Parkinson's disease, Achilles tendon).
• Hybrid word, a compound mixing Latin and Greek elements, historically frowned upon by some grammarians but common throughout the vocabulary (appendicitis, tonsillitis).
• Terminologia Anatomica, the international, Latin-language standard for naming anatomical structures, revised in 1998 from the 1895 Basel Nomina Anatomica.
• Initialism, an abbreviation pronounced letter by letter (MRI, CPR), distinct from an acronym absorbed as a spoken word (AIDS, LASIK).
• Borrowing, a term carried directly from a modern language rather than built from Greek or Latin parts (influenza, malaria, kwashiorkor).

See also

• How clinicians, lawyers, and patients read the same medical words differently: a companion piece on register and usage
• The Hippocratic tradition and the origins of Western diagnostic vocabulary
• The ongoing debate over eponyms in science and medicine, and the case for descriptive naming