Evaluating terminological oppositions

« previous post | next post »

In "Biomedical nerdview", I noted that the terms "sensitivity" and "specificity" seem to be hard even for biomedical researchers to remember, and also denote concepts that are deeply misleading from the perspective of patients and their physicians. I offered a "flash of insight" about why researchers chose to focus on the concepts — they're relevant to public health concerns, though not to patients — but I confessed to being baffled about the hard-to-remember choice of terminology. Bob Ladd responded by email:

While not wanting to take away anything from your flash of insight, I was wondering if you wanted to write another LL post, not about nerdview, but about inexcusably unmemorable terminology for related concepts that have to be sharply distinguished from one another. 

Since Bob goes on to suggest an interesting morpho-phonological theory about why some terminological oppositions are so problematic, I got his permission to post his note.

He continues:

In linguistics, my favourite candidates are (a) progressive and regressive assimilation [I always have to work out what is moving forward or backward, and when I have to talk about these I always say anticipatory and perseverative], and (b) Halliday's tonicity, tonality, and tone for what he sees as the three main functions of prosody [accent, phrasing, and contour would be more memorable].  The LL-type question that then arises is what it is about the linguistic form that makes a pair or triplet of terms unmemorable or un-mnemonic. 

On the basis of this sample of three, it's not so much the semantics of the members of the pair/set, but the phonological and morphological similarity:

sensitivity/specificity: both begin with /s/, end in -ity, and have the same number of syllables and stress pattern

tonicity, tonality, tone: all based on the root "tone", plus two of them end in -ity and have the same number of syllables and stress pattern

progressive/regressive: both are X-gressive, where X is a one-syllable prefix with limited independent semantic content.

It seems that this kind of similarity, combined with sharply distinct semantics in the specialised context, is what makes for the problem. Compare the following pairs, which are similarly specialised distinctions in a specific context, but are phonologically and morphologically distinct enough to be memorable:

medical vs surgical (in a specialised medical context)
strategy vs tactics (in a specialised military context)
agglutinating vs synthetic vs isolating (in a specialised linguistic context)

If I'm right about this, then we ought to find that anatomy students have more trouble remembering the difference between metacarpals and metatarsals than, say, the difference between the radius and the ulna or the humerus and the femur (and staying with bone pairs, radius vs. ulna should be easier than tibia vs fibula).


  1. Jonathan Mayhew said,

    September 28, 2014 @ 4:53 pm

    I nominate the scientific distinction between accuracy and precision. Both are polysyllabic, and they have similar if not identical meanings in everyday parlance. There is no intuitive way for me to remember which one is which.

  2. paul garrett said,

    September 28, 2014 @ 5:07 pm

    In mathematics, a _sequence_ is a (possibly infinite) list of numbers or other objects, while a _series_ is an infinite sum. Yes, in non-technical English, "sequence" and "series" are essentially synonyms, as far as I can tell. Nevertheless, many lower-division college mathematics textbooks attribute great importance to the distinction between these two mathematical usages, and, commensurately, some teachers of these courses take students to task about their failures to (perversely?) imbue usual-synonyms with significantly meanings.

  3. Jon said,

    September 28, 2014 @ 5:55 pm

    Hyper- and hypo- (tension, glycemia, etc). They're often indistinguishable when spoken. I'm sure some people don't realise there are two opposite conditions. High and low, or super and sub would be so much simpler and obvious in meaning.

  4. Plane said,

    September 28, 2014 @ 6:14 pm

    I've noticed a lot of people's eyes glaze over when it comes to phonetic versus phonemic.

    There's also prenominal and pronominal, perfect and perfective, precative and predicate, proverb and pro-verb, phrasal verb and phasal verb, proposition and preposition, particle and participle, to pick on 'P'.

  5. Gordon Campbell said,

    September 28, 2014 @ 6:19 pm

    Coherence and cohesion.

  6. Eric P Smith said,

    September 28, 2014 @ 6:20 pm

    @paul garrett: I agree. I am a mathematician. It's even more confusing than you suggest, because every series is a sequence, and arguably every sequence is a series. The series 1 + ½ + ¼ + … is the sequence 1, 1½, 1¾, …

  7. Ben Zimmer said,

    September 28, 2014 @ 6:22 pm

    And then there's the bane of middle-schoolers, mean, mode, and median.

  8. Tara said,

    September 28, 2014 @ 6:31 pm

    Afferent-efferent and abductor-adductor certainly pose problems for students of anatomy who are not also students of Latin. (Plus abductor versus adductor is terrible from a perceptual recoverability standpoint, which is why physical therapists repair them to A-D-ductor and A-B-ductor.) For sheer difficulty of interpretation, I think counterfeeding versus counterbleeding wins in a landslide.

  9. Brett said,

    September 28, 2014 @ 6:41 pm

    The sequence/series distinction is one that I remember seeming very strange when I first encountered it. However, through repeated use, it has become completely ingrained, to the point that "A Series of Unfortunate Events" sound distinctly wrong to me. (There's no sum, so it should be a sequence. Indeed, while every series is really a sequence, mathematical sequences are much more general than series, since a sequence can have an arbitrary set—say, the set of unfortunate events—as its range.)

  10. Nana Batrychos said,

    September 28, 2014 @ 6:42 pm

    Hypo- and hyper-. Mostly for me they have the same pronunciation. So hypochondria and hyperchondria are aurally indistinguishable (as well as semantically non-intuitive from the constituents).

  11. narmitaj said,

    September 28, 2014 @ 6:44 pm

    Hypo and Hyper might be candidates, as in hyperthermia and hypothermia, or hypertension and hypotension. In this case, I doubt there's actually much confusion when the words are written down, but there's clearly room for oral/aural misunderstanding. Especially when someone has written down the wrong word after mishearing it.

    Cosmogony and cosmology are different-but-related subjects; but less different than astronomy and astrology, which cause little problem to most but can still trip up the odd layperson.

  12. Adrian Morgan said,

    September 28, 2014 @ 7:02 pm

    In response to Ben Zimmer:

    I always used to remember that it takes a mean teacher to make you calculate the mean (since it is the most computationally intensive of the three), to which my teacher responded that it takes an average student to calculate the average.

    At the same level of schooling as when mathematics class tests students on distinguishing between mean, median and mode, geography class tested students on identifying the most common modes of transport. This provided a useful mnemonic for connecting "mode" with "most common".

  13. Jerry Friedman said,

    September 28, 2014 @ 7:13 pm

    If I'm allowed to mention it again, the way I remember "high-context" and "low-context" is that they're the opposite of what I expect.

    narmitaj: Not only that, cosmetology is very different from cosmology. And nice subtle mention of "aural" and "oral".

  14. Ralph said,

    September 28, 2014 @ 8:59 pm

    A non-technical everyday one – astrology and astronomy? No scientist would ever mix these up, but there's no obvious way to distinguish the science from the superstition for the confused layman. After all, biology, geology and ecology are perfectly respectable. I suppose the fortune tellers just had better trademark attorneys.

  15. Y said,

    September 28, 2014 @ 9:34 pm

    For linguistics, perfect-imperfect-perfective-imperfective. Proclitic vs. enclitic (why not 'postclitic'?). Accusative-antiaccusative-ergative-antiergative.

  16. Rebecca said,

    September 28, 2014 @ 9:54 pm

    I don't get why it's difficult, but I have to nominate upload and download. To me the distinction is both clear and sensible. But I hear them used backward and randomly all.the.time, so something must be going on there.

  17. D.O. said,

    September 28, 2014 @ 11:21 pm

    Thinking of "sensitivity" and "specificity"… Is "precision" and "recall" any better? They are not phonetically close at least.

  18. Keith Ivey said,

    September 28, 2014 @ 11:42 pm

    Tornado warning and tornado watch.

  19. Zizoz said,

    September 29, 2014 @ 12:33 am

    Psychologist/psychiatrist is one pair I have trouble keeping apart.

    There's also the trio of subjects geology, geography and geometry.

  20. Brett Reynolds said,

    September 29, 2014 @ 1:05 am

    One pair that comes to mind is determiner (a function in the Cambridge Grammar of the English Language but a category in the Comprehensive Grammar of the English Language)-determinative (the opposite choices).

    In general, it seems to me that linguistics makes use of more words starting with p than it really has any right to.

    There's some evidence from the second-language acquisition literature that these pairs are learned more easily if they are not studied as pair. In other words, learn one and when you've got that fairly well, then start learning the other.

  21. Thomas Rees said,

    September 29, 2014 @ 1:27 am

    Meiosis/mitosis. Both pronounced [mʌɪ-]. Why can’t the latter be pronounced with a short “i”? It’s μίτωση in Greek, not that that matters.

  22. Chris Waigl said,

    September 29, 2014 @ 1:58 am

    I probably shouldn't comment as I just admitted to my partner that I keep mixing up Joan Rivers and John Waters….

    The series/sequence distinction is one that should be just as hard in German (Folge/Reihe), but my experience agrees with a previous commenter that they became so familiar that I don't confuse them,.. In German. (It probably also helped that series/Folgen were introduced first and filled up a major math class for weeks before we added sequences/Reihen). But in English I have to remember Folge = series, and the other one is the other one.

    In Latin class, probably as an 11 or 12 year old, the nouns voluptas (lust) and voluntas (will) were introduced in the same lesson and it took me years (and learning some English where you have cognates such as voluntary and voluptuous for mnemonics. It would not surprise me that separate introductions of easy-to-confuse term pairs will help. For example, median came long after I knew about means, and I never had an issue.

    I'm surrounded by geologists these days, and they have a hard time with syncline and anticline. (So have I. It's not intuitive… or else someone has to give me the key.)

    I happen to find precision and accuracy very intuitive but my partner just tells me that in information retrieval, precision and recall (Wikipedia tells me that's the same as sensitivity, but in my corner of statistics we don't use the term) are a commonly confused pair of symmetrically opposed terms.

  23. Bathrobe said,

    September 29, 2014 @ 2:13 am

    Backwardation and contango, although the problem isn't linguistic…

  24. John Walden said,

    September 29, 2014 @ 2:33 am

    For me it's when medical tests are positive or negative. Positive results are mainly the results I don't want. So for me they are negative results. So testing positive for, let's say, high cholesterol needs a moment's thought.

  25. Nathan said,

    September 29, 2014 @ 3:23 am

    Classics: stalactite/stalagmite, and meteor/meteroid/meteorite.

    But the one that always got to me, studying social psychology, was the distinction between moderating and mediating variables. Ugh. http://psych.wisc.edu/henriques/mediator.html

  26. Mooloo said,

    September 29, 2014 @ 3:38 am

    Tidy people are frequently accused of being OCD, when it is the condition OCPD that is actually meant.

  27. bedwetter said,

    September 29, 2014 @ 3:47 am

    Translate and transcribe (in the biological sense).

    Adsorb and absorb (physical chemistry).

    Series and serial (in the context of a TV show).

    Sam Raimi and the late Harold Ramis. (For years I thought they were one person, and I could never figure out why half his films were pretty good and the other half were pretty bad.)

    Curiously though, I've never had any particular problem with sensitivity and specificity. In fact to me the terms seem pretty much intuitive.

  28. RobertL said,

    September 29, 2014 @ 6:17 am

    Metatarsal and metacarpal are pretty easy, and I say that as an ex-anatomy student. It's easy to remember that one is fingers and the other is toes. To remember which is which just remember that the word with all the "T"s is the one for the toes.

    The pair that gives me the most problems is prostate and prostrate. I always have to stop and think about them.

  29. flow said,

    September 29, 2014 @ 6:18 am

    oh yeah, transcription—transliteration (and why you can do both with Russian/English, but only one of those with Chinese/English).

    affluent—effluent. affect—effect. it's—its, then—than (maybe not belonging here, but still). anti…—ante…

  30. Rodger C said,

    September 29, 2014 @ 6:53 am

    In chemistry there's molarity and molality, one of the reasons I became an English major.

  31. Adrian Bailey said,

    September 29, 2014 @ 7:18 am

    Continuing on John Walden's theme, the terms positive/negative feedback are problematic.

  32. BlueLoom said,

    September 29, 2014 @ 7:22 am


    On a trip thru Luray Caverns in Virginia when I was perhaps 9 or 10 yrs old (I'm over 70 now), a guide said, "Always remember that a stalacTite comes from the Top." I may have to run thru the mnemonic in my mind quickly before I use either one, but it tends to keep me from making a total fool of myself.

    Now, about horizontal/vertical…I always have to look to see which one lines up w/ the horizon before I use either of them, spoken or written.

  33. bedwetter said,

    September 29, 2014 @ 9:00 am

    How do people remember if the wide blade on a standard NEMA electrical plug is live or neutral?

  34. Adrian Morgan said,

    September 29, 2014 @ 9:54 am


    I just thought of a mnemonic for "stalactite" that might, perhaps, work for some people (personally, I've never needed one).

    First, observe that people typically say "stalactites and stalagmites", not "stalagmites and stalactites". (Actually, ngrams data puts the ratio at a modest 1:3, which surprises me, as I find the latter most unnatural.)

    Now, consider a drop of water as it falls from the stalactite to the stalagmite. The falling drop encounters the two formations in the same order that they are typically uttered. Convenient!

    (Or maybe you prefer the oft-used mnemonic that involves tights — the article of clothing — dangling from the ceiling. So be it: my suggestion just adds another possibility to choose from.)

  35. Ralph said,

    September 29, 2014 @ 10:38 am

    "I may have to run thru the mnemonic in my mind quickly before I use either one, but it tends to keep me from making a total fool of myself."

    Um, how exactly?

    I'm wondering if anything in the entire history of the world, however trivial, ever hinged upon distinguishing between stalactites and stalagmites correctly? Did "Careful, don't sit on that stalactite!" once have hilarious consequences?

    Perhaps troglodytes have hundreds of words for pointy limestone deposits, but I'm not convinced that we surface dwellers even need two.

  36. Pflaumbaum said,

    September 29, 2014 @ 10:47 am

    Meiosis/mitosis, stalactite/stalagmite and phonetic/phonemic are great calls. I love the fact that the latter are a minimal pair when written. Genetic/genomic is similar.

    The ones in linguistics I find hardest are the aspectual terms achievement and accomplishment. Having activity and atelic hovering round the scene doesn't help massively either.

  37. Chris Waigl said,

    September 29, 2014 @ 10:51 am

    Prone and supine.

    I need a mnemonic for stalactite/-mite, too, but the one I learnt back in German (the words are nearly the same) uses some taboo language.

    Another one for remedies: I hear much complaining about ana-/cata- pairs like anaphoric/cataphoric in linguistics or anode/cathode in physics/chemistry, but having been made very familiar with the Greek ἀνά/κατά (up and down) I rarely have to stop and think.

  38. Pflaumbaum said,

    September 29, 2014 @ 10:51 am

    Also, pleistocene/pliocene/miocene. Nightmare.

    And there's one nagging at my mind, but I can't think of the second term. It either goes with deictic in linguistics, or with diegetic in narrative studies. But I think that usually contrasts with mimetic, which seems unproblematic.

  39. Vic said,

    September 29, 2014 @ 10:52 am

    When I was 9, the guide at Crystal Cave, PA said stalaCtites were on the Ceiling, and StalaGmites were on the Ground. That was 1957, and I never forgot.

  40. Ralph said,

    September 29, 2014 @ 11:21 am

    "Translate and transcribe (in the biological sense)"

    To transcribe is to copy out; to translate is to change languages. This one makes sense and is easier to remember if you understand the molecular biology.

    The genetic code is recorded in very long unwieldy molecules of DNA, called chromosomes. The language is nucleotide bases. When a cell needs to read the code to make a protein, it first makes a direct copy of the relevant part of the chromosomal DNA. The copied section is a much smaller and more mobile molecule called a messenger RNA. RNA is written in the same language as DNA, nucleotide bases. DNA > RNA is just direct copying, so it's called transcription.

    In order to make a protein, the cell now reads along the messenger RNA sequence to construct the amino acid sequence that constitutes the protein. However, there is no direct chemical correspondence between the nucleotide bases of DNA/RNA and the amino acids of proteins. DNA/RNA are written in one langauge (nucleotide bases), proteins are written in a different language (amino acids). So this is called translation.

  41. Jerry Friedman said,

    September 29, 2014 @ 12:04 pm

    Rebecca: Upload and download brings us to backslash and the retronym forward slash.

    I always have to think about what putting the date of a meeting back or forward means.

  42. Ralph said,

    September 29, 2014 @ 12:23 pm

    Regarding upload and download… now that we're uploading and downloading to and from the cloud, it all makes much more sense.

  43. David L said,

    September 29, 2014 @ 12:32 pm

    Stalactites and stalagmites: a long time ago I learned the mnemonic "the mites go up and the tights come down." This was highly amusing to 14-year-old boys.

  44. Jerry Friedman said,

    September 29, 2014 @ 12:50 pm

    Brett Reynolds: I keep forgetting to add that I'm not surprised it works better if you don't learn them as a pair. As far as I know, no one has a problem with accelerate and decelerate or with associate and dissociate. The familiarity of de- and dis- must help too. However, I think it would be hard to teach anatomy students afferent and efferent separately.

  45. Bob Ladd said,

    September 29, 2014 @ 12:53 pm

    Thanks for all the examples, everyone. Many of them certainly fit the suggestion I made to MYL, but not all of them, which shows that more is going on here than just phonetic similarity. There's also the fact that some of these seem to cause problems for lots of people (like stalactite and stalagmite) while others are more idiosyncratic (like I've never even thought of hypo- and hyper- as a problem, though it would seem to fit the pattern I suggested to Mark).

    It's also nice to see that some of the mnemonics people report for tricky pairs are ever weirder and more far-fetched than my own.

  46. J. W. Brewer said,

    September 29, 2014 @ 2:04 pm

    You can find plenty of hits Out There for, e.g., "hyperdermic needle" — the problem, I think, is that "hyper-" is widely understood as a freefloating morpheme, meaning something close to "super-" (maybe in non-technical coinages something like "even more super than super"?) whereas "hypo-" is not understood at all (and is certainly not available for new coinages) outside of specialized circles, so it's easy for it to get muddled with the very-similar-sounding prefix people do know. (Plus specific uses of hypo- people do know are likely etymologically opaque – e.g., most people who know the word "hypodermic" do not have the understanding that it = "under-[the]-skin.") This is slightly different from situations where people know both words in the pair but can't keep them straight – much hyper/hypo confusion reflects imho lack of awareness that there are in fact two different morphemes in the first place.

  47. Chris Henrich said,

    September 29, 2014 @ 3:15 pm

    Here is another math-related lament: the letters m and n. How many ways can they be confused? They look alike. They sound alike. They are next to each other in the alphabet, and on the standard keyboard. And we mathematicians are fatally fond of using them together. With similar, but importantly different roles. At least one book I recently read seemed to switch them around at times, when quoting one theorem in the proof of another. As if math wasn't hard enough anyhow.

  48. Y said,

    September 29, 2014 @ 4:11 pm

    hypo- and hyper- are extra fun when you have a professor who speaks with a non-rhotic accent, but hypercorrects every once in a while. True story.

  49. hector said,

    September 29, 2014 @ 4:22 pm

    "inexcusably unmemorable terminology for related concepts that have to be sharply distinguished from one another. "

    Is the ur-example of this problem the distinction between right and left? Most people occasionally use the wrong word. Both are words of one syllable, thus easy to say, and often need to be said as quick responses to events. If one of the two was one-syllable, and the other multi-syllable, would this reduce the number of mistakes?

  50. Belial said,

    September 29, 2014 @ 5:22 pm

    the letters m and n. How many ways can they be confused? They look alike. They sound alike.

    …and what is more they hate each other very much.

  51. bedwetter said,

    September 29, 2014 @ 5:47 pm

    @Ralph, re transcribe/translate — that's going to be enormously helpful. Thank you!

    I once read — possibly in an Enid Blyton Famous Five adventure? — that stalactites have to hold on tight or they'd fall.

    'Anode' has the word 'no' inside it, so – ha! — it's positive! (Many of my home-grown mnemonics are perverse.)

    There's also yin/yang. Which is not helped by lingam/yoni, although for some reason I have no problem remembering which is which in the latter.

  52. narmitaj said,

    September 29, 2014 @ 5:53 pm

    @ hector "the distinction between right and left?"

    As well as normal confusion (I got confused during my early driving lessons and even briefly – and non-disastrously – during my test at one point), there is also the problem that crops up in film-making and other environments when people are facing each other and trying to give and take directions. The cinematographer might say to the actress "move to the right" and the actress moves to her right, and the cinematographer says "no, the other right", his right, camera right or screen right.

    Of course everyone would be on the same script page if they could say "move two steps to the west", but it is not really practical, especially indoors on cloudy days. Besides, "east" and "west" (est-ouest) are probably confusable too.

  53. Brett said,

    September 29, 2014 @ 7:06 pm

    @Vic: I've never had trouble remembering the difference between stalactites and stalagmites. What your mnemonic ensures I remember is that the word "stalactite" has the letter "c" in it. Until I heard the mnemonic from a cave tour guide, I had no idea that the spelling was not "stalagtite." Very few people seem to distinguish the voicing between the "c" of "stalactite" and the "g" of "stalagmite."

    @narmitaj: People often tend to sneer at the nautical use of "port" and "starboard," as if the terms were redundant. However, the point of the terms is that they are not equivalent to "left" and "right"; "left" and "right" are tied to a person's orientation and thus not consistent between individuals facing different directions, while "port" and "starboard" are unambiguously measured relative to a vessel.

  54. Adrian Morgan said,

    September 29, 2014 @ 7:36 pm

    Re left and right:

    One thing that helps is that I've never had any difficulty remembering (as a propositional fact) that in western cultures we read from left to right. So I can tell them apart by considering the direction in which I read.

    But one thing that has sometimes tripped me up is that, when you turn left, it is your RIGHT hand that swings around while your left hand is a motionless pivot, and when you turn right, it is your LEFT hand that does all the moving (and is therefore in a sense more salient to the act of turning). So by fixating too much on the hands rather than the directions, you can get them muddled.

    A related question is how to interpret left and right on a computer screen. It is conventionally interpreted from the user's perspective, but that is something each of us has to learn; it is by no means self-evident. After all, a monitor is metaphorically like a face, so it could just as easily be interpreted relative to the computer.

  55. Y said,

    September 29, 2014 @ 8:32 pm

    I knew someone who also got confused by left and right, but not by port and starboard. It was easier for her to have driving directions that way. She'd sailed some, but didn't grow up on a boat or anything like that.

  56. narmitaj said,

    September 30, 2014 @ 2:50 am

    Port and starboard are probably easier to keep track of than left and right due to the different-sounding words and to the word order – no-one says "starboard and port", whereas right and left are more interchangeable.

    P.S. Also, I think of P.S.

  57. AllenV said,

    September 30, 2014 @ 5:51 am

    What's more confusing is the concept of a cathode and anode (which some of you have already mentioned), which during discharge in a cell are the + and – electrodes respectively, but switch during recharging so that the cathode is – and the anode is +.

  58. David J. Littleboy said,

    September 30, 2014 @ 6:59 am

    Jerry F. said: … "is that they're the opposite of what I expect."

    This is a really bad idea. It works fine for a while, and then you get familiar with the terms and comfortable using them with their correct meanings, and you can't remember which is the wrong one you used to expect.

    In Japanese, there's 緑 and 縁. Those two are _always_ on the test, but never matter in real life, because context will always disambiguate for you.

  59. J.W. Brewer said,

    September 30, 2014 @ 7:44 am

    hector: there are certainly languages out there where the words for "left" and "right" have different numbers of syllables (e.g. Japanese "hidari' v. "migi," or Italian "sinistra" v. "destra"), so the study in cross-linguistic variation in getting the two muddled simply needs someone to write up the grant proposal.

  60. D.O. said,

    September 30, 2014 @ 10:19 am

    I think, we are brainstorming the wrong problem. There is plenty confusion and ambiguity in language. Especially when lay people use technical vocabulary. The question is when experts who work with these terms professionally get confused.

  61. Bob Ladd said,

    September 30, 2014 @ 1:24 pm

    @D.O. But we're not just talking about ordinary confusion – most of these examples are about terms for things or concepts that are direct opposites or otherwise matched pairs or sets (like mean, median, mode). My original suggestion was that such terms are harder to tell apart if they're phonologically and/or morphologically similar. Some of the examples certainly bear this out – I particularly like @bedwetter's examples of yin/yang (hard to remember) and lingam/yoni (easy to remember). (At least, I agree with him/her about which pair is easy and which is hard.)

    Also, we're not just talking about experts and technical terms. During the period in my life when it might have made sense for me to learn some Lithuanian, I never managed to get straight which of the pair šiltas/šaltas means 'warm' and which means 'cold'.

  62. D.O. said,

    September 30, 2014 @ 3:04 pm

    Direct opposites, I would call them not. Mean, median and mode are what is known generally as "central tendency" and my intuition is that most of the confusion is because concepts (and not words) are so close. In the yin/yang pair, for example, I do not remember which is which, but I hear/read about them maybe once a year and always as a pair indicating two opposites, not any of their specific characteristics.
    sensitivity/specificity are somewhat similar in sound (not for me, sen- cannot be confused with sp- cluster), but they probably are confused because they describe analogous things.
    tonicity, tonality, tone: I've no idea what tonicity is (neither apparently the WordPress spellchecker), but if I confuse them that's because I don't know them.
    progressive/regressive: How many people make mistake of them as political terms (OK, regressive is not actually used as much)? Once again, most probably the source of confusion is simply inactive knowledge.

    To sum up. If we want to distinguish phonological (do I use this word correctly?) similarities as a source of confusion we should eliminate a large effect of people generally having somewhat blurred mental image of words they rarely use.

    P.S. šiltas/šaltas. English pair hot/cold is not the worlds apart either.

  63. Rubrick said,

    September 30, 2014 @ 4:09 pm

    Inter- and intra- anything are persistent troublemakers.

    Also, while not really jargon (I think), I have a terrible time with turgid vs. turbid.

  64. Peter Erwin said,

    September 30, 2014 @ 4:23 pm

    Port and starboard are probably easier to keep track of than left and right due to the different-sounding words …

    And that seems to be the reason that "larboard" was replaced by "port", because "larboard" and "starboard" are (or were) a pair that lends itself to confusion.

  65. Peter Erwin said,

    September 30, 2014 @ 4:27 pm

    You can find plenty of hits Out There for, e.g., "hyperdermic needle" — the problem, I think, is that "hyper-" is widely understood as a freefloating morpheme, meaning something close to "super-" (maybe in non-technical coinages something like "even more super than super"?)

    And even in techincal coinages: a "hypernova" is a stellar explosion which is significantly more energetic than a "supernova".

  66. Eric P Smith said,

    September 30, 2014 @ 4:45 pm

    @Adrian Morgan: actually, for populations of more than a very few members, the mean is the least computationally intensive. To compute the median or the mode, you need to either keep track of every member, or use quite a complicated optimisation. To compute the mean, you only need to keep track of the count and the running total.

  67. narmitaj said,

    September 30, 2014 @ 5:21 pm

    There's also altitude/attitude in aviation circles, one being simple height above ground (or sea level), the other being orientation in three dimensions (pitch, roll and yaw).

    While on aviation matters, I've been reading up a bit recently on the Apollo programme and it has reminded me of instances of a related issue, where opposite actions are called for using, in part, the very same word. For instance, Go and No Go. So the flight director (eg Gene Kranz) asks all his flight controllers for a Go/NoGo for powered descent to the lunar surface; you'd think there would be room for confusion, especially over radio circuits where the No might be obscured. (A similar situation – radios, and in this case casual use of the term "take-off" – cropped up in the disastrous Tenerife 747s collision in 1977).

  68. Ralph said,

    September 30, 2014 @ 6:34 pm

    I'm astonished to hear about Go / No Go. Perhaps there were other context requirements for a positive clearance?

    For just that reason, in modern aviation radio communication, "take-off" is only ever used in conjuction with a clearance to do so. In other contexts, "departure" is used. Thus, for example:

    "Speedbird 123, you are number 3 for departure"
    "Speedbird 123, you are cleared for immediate take-off"


  69. Alex Fink said,

    September 30, 2014 @ 7:04 pm

    Sai's favourite example, as it were, of this comes from an artificial intelligence course he took from John Searle as a Berkeley undergrad. Searle distinguished intentionality and intensionality (with relish), identically pronounced, though not exactly a semantic minimal pair. A system has intentionality if it can intend to do things; it has intensionality if it has intensions i.e. internal concepts which are signified by the words it uses.

  70. DCA said,

    September 30, 2014 @ 7:32 pm

    I'll offer the statistics triplet unbiased, efficient, and sufficient. All have common meanings which match only (perhaps) for the first. I tell my students that statistics has suffered from Karl Pearson's approach of never using English when Latin would do (eg heteroscedasticity) and R. A. Fisher's preference for (instead) using common English words in new senses.

  71. rosie said,

    October 1, 2014 @ 3:39 am

    @Jerry Friedman Re back/forwards. Usage of those confuses me sometimes because people usually use them what seems to me to be the wrong way round: "back" meaning "further into the future", which I think of as "forwards".

    I'm not an anatomist but "abductor" and "adductor" don't faze me because the construction of the words shows that the ab… leads away and the ad… leads towards.

    By contrast it's not obvious to me which of the words subjective/objective has which meaning. The words don't give me any clue as to which means "a matter of opinion" and which means "a matter of fact".

    Two more pairs: cyclone/anticyclone and veering/backing (w.r.t. wind changing direction).

    And is a male-to-male connector a connector whose ends are male and which connects a female socket to another female one, or vice versa?

    And given that e.g. northwards means towards the north, it was always confusing to me that a northerly wind is one that blows /from/ the north, southwards.

  72. Rodger C said,

    October 1, 2014 @ 6:54 am

    And given that e.g. northwards means towards the north, it was always confusing to me that a northerly wind is one that blows /from/ the north, southwards.

    That's because the North Wind was originally pictured as a guy standing in the north and blowing.

  73. iching said,

    October 1, 2014 @ 8:36 am

    "Effectiveness" – "Doing the right thing"
    "Efficiency" – "Doing the thing right"

  74. Brett said,

    October 1, 2014 @ 8:41 am

    @Rodger C: I found the naming of directional winds odd until I realized that was the original metaphor. Since then I've not had trouble.

  75. iching said,

    October 1, 2014 @ 8:50 am

    "Induction" – "Reasoning from particular to general"
    "Deduction" – "Reasoning from general to particular"

  76. Ralph said,

    October 1, 2014 @ 9:04 am

    "I found the naming of directional winds odd until I realized that was the original metaphor"

    The convention for wind direction is not arbitrarily derived from a metaphor. We care where the air mass has come FROM because that determines it's properties. Air from the north is cold; air from the ocean is moist.

  77. iching said,

    October 1, 2014 @ 9:05 am

    "diffusion" – "effusion"

  78. iching said,

    October 1, 2014 @ 9:20 am

    différence – différance (Derrida's neologism)

  79. Bob Ladd said,

    October 1, 2014 @ 10:55 am

    About wind directions: what would you do if you were a designer of little icons for a weather forecasting site? On the BBC weather site, the wind icon is a circle with a number on it (for the wind speed) and an arrow sticking out from the rim of the circle (for the wind direction). If the wind is from the west, the arrow points east. This directly represents what the wind is doing (it's iconic in the strict sense of the word), but it's still potentially confusing if you are expecting the arrow to point to the compass direction that the wind is named after. There is a mouseover message to assist the potentially confused.

  80. Ralph said,

    October 1, 2014 @ 11:30 am

    Old fashioned mechanical weather vanes have a full arrow, with a head at one end and feathers at the other, and the head points to where the wind is coming FROM. Similarly, technical weather maps that meterologists use follow this convention, for example here:


    The lines protruding from the circles on the page above show where the wind is coming FROM.

    But, as you point out, informal weather forecast symbols are utterly confusing. The BBC site, and many others, have a circle with an arrowhead protruding, showing where the wind is blowing TO.

    I think the best solution for the BBC would be to have a single line protruding from the circle showing pointing to where the wind is coming FROM (consistent with prior convention) but with an arrowhead on the line pointing IN toward the center of the circle, to clarify matters for those who are unaware of the convention. I think this would be unambiguous.

  81. Mark Stephenson said,

    October 2, 2014 @ 8:29 pm

    I've never had any trouble with stalactites and stalagmites since my father taught me 'Stalactites have to hang on tight, and stalagmites might reach them if they grow far enough."

RSS feed for comments on this post