Richard and I were working on the Bayesian Workflow book, I had an idea for an example to include and volunteered to write it up, and Richard suggested I “take a pass” on it. Then we realized that “take a pass” has two opposite meanings: “Do it” and “Skip it.” Which reminded me of something we discussed several years ago, that “sanction” is one of those funny words with two nearly completely opposite definitions. According to dictionary.com, the first definition of this verb is “to authorize, approve, or allow.” But “sanction” also is used in the opposite sense, to mean “punish.” Similarly, the noun can either mean “permission” or “penalty.” It can get confusing at times, to read that some behavior is being sanctioned.
Richard said that in German there are many such words with opposite meanings; they are called Janus [two-facing] words. In English they are called contronyms; here are wikipedia and Merriam-Webster, which includes a fascinating story of the derivation of the word “fast.” The wikipedia entry for Contronym is in 13 languages, including German but not French.
Isn’t it a matter of the preposition? “Take a pass at” = “give it a try.” “Take a pass on” = “Skip.”
Dmitri:
Could be. The funny thing is, I now don’t remember that “take a pass” conversation at all—I guess it happened several months ago—and if you would ask me now what is meant by “take a pass,” I don’t know how I’d answer. Not because of the contronym thing but just because it’s not an expression I would usually use or hear.
So maybe this is less of a story of a contronym and more of a story of someone using an expression with which I was unfamiliar, followed by my attempt to make sense of it. The episode was all worth it for me because then I learned about the cool expression “Janus word.”
I think it’s idiomatic even without the preposition:
“Do you want a helping of jello salad?” “No, I’ll take a pass.”
“Can somebody write the intro to this paper?” “Sure, I’ll take a pass.”
I’d probably say:
“Do you want a helping of jello salad?” “No, I’ll take a pass.”
“Can someone write the intro to this paper?” “Sure, I’ll take a pass at it.”
In my mind, the first one means something like “I’ll request an excuse note” obviously not literally, but basically ask for permission not to participate.
My impression is “take a pass at” ie. to attempt is maybe derived from something like jousting or some other sporty thing, like ring tosses off a merry go round, you literally “go physically past something” and “try to do the action one is supposed to do when going past the thing”.
According to wiktionary, the French verb sanctionner has tye same contradictory meanings as in English.
The French word “plus” is my particular nemesis. It’s meaning is opposite depending on how it’s pronounced. Somehow French people can tell which pronunciation by looking at it.
Cleave.
The idea that “take a pass” is used in the “I will do it” or “I will try it” sense is new to me. In my world it has always meant bypass, skip, forego, exclude etc something. The only exception that comes to mind – not sure if I’ve heard it used or not – is to physically pass something, as in “swing by”, e.g., “I’ll take a pass by Home Depot and see if they’re still open so I can get that Staples-brand paper shredder”
I have also never heard “take a pass” to mean “I will try it”, but my wife has. Maybe this means that people were always telling me that they’d do things and I was assuming they wouldn’t. I should upgrade my opinion of humanity.
“I’ll take a pass at it” is something a musician would say.
I don’t understand a world in which this silly language post gets 6 comments right away, while yesterday’s thoughtful post on prediction markets only got 1 comment in total!
It’s easy to read a short post and toss in your 2 cents about contronyms. Reading and responding to a big post on a complex topic takes more work.
Anoneuoid took a pass at it. But so did I (alternative meaning).
I visit the blog for statistics posts; prediction markets do not concern me. I’m sure other readers visit for their particular interests among the wide range of topics covered here. But all readers can read and comment on “silly” posts, so the proportion of readers disposed to read and comment on them is probably quite high.
Andrew:
Your observation suggests a cool example of why the results of a single research project, however apparently strong at first glance, need to be confirmed with further research. It also suggests an interesting teaching opportunity.
Your apparent surprise seems to come from a belief or idea that the quality of the writing and/or insight on a given post should be the primary causal factor in the number of comments – e.g., that the quality of the post should be measurable via the number of comments. However, your observations seem to contradict this idea and instead suggest just the opposite: lightweight posts with little serious content get the most comments.
So, which is true? Does great writing and insight drive the number of comments, or just the opposite? I suggest you pose that question to students and assign them the task of developing the research methodology to determine the answer. The data is everything you can provide about your blog – everything that’s publically available, plus whatever additional data you might have like the number of hits. Provide as few constraints as possible on the assignment, leaving the field wide open for students to determine what constitutes a “quality” post and how to measure how readers interact with posts of different quality. This would be a group project, so maybe groups of 3-4 students in each group and ideally five or more groups to get the greatest breadth of possible approaches. Ideally you could pursue a time-intensive option and have the students actually write the code and run their methodology and present conclusions then have a class discussion and critique on each method. How many different approaches were there? How did the different approaches drive the conclusions? But you could shorten it to one class period, skip actually coding and analysis, and have each group present an outline their methodology to the class followed by critiques or Q&A regarding the strengths and weaknesses of each methodology.
The point of the exercise is that even generating a methodology gets sticky fast. How do you measure a “deep and insightful” post, vs. one that is “light and entertaining”? Is a 2-D linear (e.g., “deep” vs. “light”) scale appropriate? Or are more dimensions/axes/variables required? How many groups stick with the simple 2-D scale? What are their justifications? An obvious mitigating factor is that there needs to be some method of controlling – apologies, adjusting! – for the number of hits on a given post. Also, as with yesterday’s post, there seems to be an effect of competition between current and recent posts. Can the competitive effect be scored? Is the number of hits or comments a good measure of the attention people give to the post? Jessica often writes long and complex posts that don’t get a lot of comments. But OTOH, they require a lot of careful thought to understand, so I can easily imagine that I spend the most time on her posts even if I don’t comment on them. So the lack of comments on a long post doesn’t necessarily equate to less time spent.
OK, now I spent too much time on this! But it would be a cool assignment. It’s not really a stats assignment. It’s a science assignment. It’s about developing a realistic concept about the number of variables influencing an outcome and trying to develop a realistic concept of which variables to measure and how to measure them to determine causal effects. IMO that’s the big challenge in social sciences. Once that is controlled to some reasonable degree, the stats are a technical matter.
Chimpunk:
Agreed that this is a fun topic, not just regarding this job but regarding media in general. There are lots of measures, including clicks, reads, comments, and—the hardest thing to measure—impact in people’s lives. Back when I was a kid, even low-rated network TV shows would get tens of millions of viewers in each episode, but with rare exceptions I assume the individual episodes had essentially no more impact than that of the Wonder Bread we ate in our lunches every day. Movies (with the exception of the occasional blockbuster) had much lower viewer numbers, but the audience was paying attention.
Regarding our blog overall: I don’t know how many readers each post gets or even how many readers we have overall. I think the blog software has some tabulation features but I purposely never set them up because I don’t want to waste my time looking at those sorts of numbers.
Regarding my comment above: I’m not shocked that a fun post about linguistics would get more comments than a technical post about forecasts and markets—especially because my conclusions are moderate, relative to the space of discussions about the topic. I was still a little bit surprised, though, because political markets are currently a newsworthy topic, and I thought some commenters would have opinions to share.
“I purposely never set them up because I don’t want to waste my time looking at those sorts of numbers”
That’s amusing coming from a statistician, who might otherwise be trying to sell everyone on the value of analyzing blog stats. But its the right choice: it’s more fun, and creates more value, to focus on the work itself than on the indirect analysis of it.
Interesting to see the variety of impressions people have regarding “take a pass”. Chipmunk says that using it in the sense of doing something is new to them, whereas for me that is the primary meaning.
I’d be curious if there were demographic differences. I’m in my 30s and I’d assume someone “taking a pass” on something to mean taking a shot, whereas if someone weren’t doing something they’d “pass on it”.
In my 30s too (Texas). It’s not a phrase I’d use, in either meaning, and not one I really hear – but I’ve seen it in writing. I’d tend to assume “I’ll take a pass” to mean more or less “I’d rather not”, but “I’ll take a pass at it” to mean “I’ll try it but the result may not be any good”.
“I’ll pass” is much more normal to me than “I’ll take a pass” to mean “I’d rather not”. I would have assumed “I’ll pass” was short for “I’ll pass that up”, though.
The idiom used to be “*make* a pass at”, not “*take” a pass at”, as I remember it from very long ago. Nowadays, I would say “I’ll take a pass [on]” if I want to skip the activity. To me, that means much the same as taking a “raincheck”. IOW, postpone until later (which may never come).
Doesn’t “make a pass at (someone)” mean something different, more or less equivalent to “ask (someone) out” (maybe with a bit more sexual than just dating intent though)?
No one has yet mentioned either “rent” or “peruse” which I would have thought are the most famous common contronyms in English. As an interesting sidelight, my spellchecker flags the spelling of the word, contronym! Anyone else have a spellchecker which refuses to recognize the word as legitimate American English?
My favorite contronym expression is “table a motion” because in the U.S. it means to kill it, while in the UK it means to bring it up for discussion.
https://tips.slaw.ca/2021/research/lets-table-that/
A history teacher once told my class that the two meanings of ‘table’ nearly delayed the D-Day invasion, because when Eisenhower suggested to Montgomery that it was time to consider invading somewhere in France, Monty said “let’s table that” and Eisenhower thought he was being rebuffed.
As far as I know, there is no support anywhere for this anecdote.
I forgot to mention that “table a motion” in Canadian English is identical to that of British English and thus opposite to that of American English.
Andrew, when you say Richard do you mean Richard McElreath? If so, exciting news!
Kiran:
Yes, it’s this book.
“But all readers can read and comment on “silly” posts, so the proportion of readers disposed to read and comment on them is probably quite high.”
Speaking of silly, that word is a diachronic contronym. Historically, it meant the opposite of its modern meaning–it meant serious.
The word, “silly” has been much studied, including Shakespeare’s Juliet being called a silly girl, as in blessed/innocent, because she is a child and hence, silly. Or so the story was told to me.
https://talesoftimesforgotten.com/2016/11/28/the-facinating-evolution-of-the-word-silly/#:~:text=The%20word%20silly%20is%20ultimately%20derived%20from%20the,ges%C3%A6lig.%20Originally%2C%20this%20word%20meant%20%E2%80%9Cblessed%E2%80%9D%20or%20%E2%80%9Cfortunate.%E2%80%9D
In my idiolect:
I’ll pass / I’ll pass on that = I won’t participate
She made a pass at me = she hit on me
I’ll take a pass at that = I will try [but I don’t use this myself, only understand it passively]
Is “dropped” also a contronym?
“Imply” means “suggest” but it also means “necessitate.” Does correlation imply causation?
An annoying near-example that I find occurs infrequently enough that I mix up their meanings is being “thrifty” as opposed to being a “spendthrift”.
Sometimes the context helps, but often I find I disagree with the author’s prescription of whether one should be more or less profligate with one’s money!