I'm Mike Pope. I live in the Seattle area. I've been a technical writer and editor for over 30 years. I'm interested in software, language, music, movies, books, motorcycles, travel, and ... well, lots of stuff.

Read more ...

Blog Search

(Supports AND)

Google Ads


Subscribe to the RSS feed for this blog.

See this post for info on full versus truncated feeds.


Life would be dull without mistakes.

— Oscar Wilde


<October 2020>



Contact Me

Email me

Blog Statistics

First entry - 6/27/2003
Most recent entry - 10/11/2020

Posts - 2629
Comments - 2638
Hits - 2,313,917

Entries/day - 0.42
Comments/entry - 1.00
Hits/day - 366

Updated every 30 minutes. Last: 11:36 PM Pacific

  10:26 PM

Covid and name-spelling

I've always had an issue in trying to tell people what my last name is. As I've written about before, people have a hard time understanding the name Pope. "Hope?" "Polk"? There's just something about that combination of consonants that's always hard to hear, especially over the phone.

Well, Covid certainly hasn't helped. I'm not a particularly clear talker under the best of circumstances. Add to that wearing a mask and talking through a plastic shield, and I feel like I practically have to yell to transact business when I'm at a store. One of the local shops posted the following sign, which might seem vaguely rude, but I think everyone understands the issue.

However, even yelling is often not enough to make my name clear to people across the counter. For some people, it can help to spell their name. But the letters P-O-P-E sound similar to other letters, like B, D, E, and spelling my name can just generate a second round of "did you mean …"? Another option is to say something like P as in Peter, and so on. When I talk on the phone with people who are accustomed to the problems of transcribing names, they'll sometimes do this.

Long ago the military solved this problem by inventing a so-called spelling or phonetic alphabet. You probably know this alphabet, which starts with Alfa (or Alpha), Bravo, Charlie, and finishes with Yankee, Zulu:

The beauty of this particular alphabet is that the names were carefully chosen to be unambiguous. Although the letters B and P sound similar, the words Bravo and Papa don't, so there's not much chance of them being confused, or Delta and Golf, or Mike and November. If you look through this alphabet, you'll see that none of the names sound like any of the other names.

So now when I am masked up and talking through a shield, I've taken to preemptively spelling out my name using this alphabet. I'll go to the pharmacy, and when they ask what name the prescription is under, I'll say "POPE, Papa-Oscar-Papa-Echo!" I sometimes get funny looks after this, but so far, no one has misunderstood what I'm spelling out.[1]

[1] Many people who were in the military know this alphabet, and a guy I used to work with who was ex-military taught me the term Charlie Foxtrot as a way in polite company to refer to a clusterfuck.

[categories]   ,


  07:30 PM

As I explained in part 1 and part 2, Duolingo's pattern- and repetition-based instruction is supposed to teach you a language without grammar instruction. How's that working for me?

I should note that I've had a lot of language instruction—I rassled with two-way prepositions in German and with the subjunctive in Spanish. I can explain phrasal verbs in English. I know what vocative means in "vocative comma."

So I'm probably an outlier in Duolingo's audience. Still, language learning is language learning, and even when you're studying a language the traditional way, you're only going to get better if you practice and practice, which is one of Duolingo's mantras. Could I learn Latin through repetition alone and with no explanations?

Possibly, but I wasn't patient enough to find out. Even at the very beginning, when I was presented with a new pattern, I was both learning it and analyzing it. To repeat an example from earlier:

Feminae domi sunt. (The women are at home)
Viri domi sunt. (The men are at home)
Pueri domi sunt. (The boys are at home)
Puellae domi sunt. (The girls are at home)

From this I abstracted that one plural ending was -i and another was -ae. I might have even concluded that -i is a plural for masculine nouns and -ae for feminine ones. (I guess I also abstracted the idea of noun genders and singular/plural verb conjugations, none of which Duolingo has breathed a word about.) So I'm not just learning the patterns, I'm starting to fill in a chart in my head of noun declensions and verb conjugations.

The first thing that threw me a bit was encountering this contrast:

Femina domi est. (The woman is at home)
Estne femina domi? (Is the woman at home?)

My previous language experience had not primed me to expect that when a verb is used as a question, it takes (or can take) an -ne suffix. Because this was novel to me, I stumbled over it repeatedly until—just as Duolingo hopes—I just went with it. (Mostly; I still forget sometimes.)

As I progressed, I kept mental tabs like this on grammatical aspects as they were introduced. For example, from sentences like this[1]:

… I tucked away that the 2nd singular form of verbs (present tense) ends in -s and that in the singular, direct objects (i.e., accusative case) have an -m at the end.

Eventually I got to something that I just could not figure out. Compare these sentences:

ego litteras latinas lego. (I read Latin literature [lit. "Latin letters," how poetic])
ego litteris latinis studeo. (I study Latin literature)

Here's where my need to analyze the grammar worked against me. In both sentences, "Latin letters" is the direct object. But which is the form for the direct object, litteras or litteris? I kept seeing both, it seemed to me, and this made no sense to me.

Maybe I could have just accepted this contrast, the way a Roman toddler would have eventually gotten it through sheer repetition. But after I'd already spent some time deducing that the -as ending was an accusative (direct object) ending, what was I to make of this seemingly arbitrary difference?

On every exercise, when you've submitted your answer, there's a Discuss link at the bottom:

When I was sufficiently flummoxed by this direct-object thing, I clicked the link. Sure enough, I was hardly the only one to have had this question. Someone who knew Latin had explained: the verb studere ("to study") takes a dative object. That's why it's ego litteris latinis studeo.

This is where Duolingo's philosophy and traditional instruction really part ways. The term "dative object" instantly cleared this up for me; I'd encountered this in German and in Old English. I imagine that the folks at Duolingo assume that the term "dative object" would be gibberish for many (most?) of their students, so of course they don't formally provide an explanation. But they've provided the discussion forum as a backdoor, so to speak, where learners can talk to one another and where you can often get the grammatical explanations you won't get directly from Duolingo.

At this point, I decided I wanted to supplement Duolingo. This was a learning experience in itself. The teaching of Latin was the template for language instruction for, what, the last 15 centuries? And boy, a lot of Latin coursework is the epitome of old-skool (haha) language learning. The classic approach, it seems, is that after a quick lesson on how to pronounce the letters, you learn all 6 cases for the first noun declension! Fun times.

After looking around and reading a lot of reviews, I settled on three books:

And I've got a notebook and I even went and got 3x5 cards to make flashcards with. Not to mention that I can find stuff on the internet.

I'll keep doing Duolingo, because I do actually have faith in the principles they're following. Daily practice and repetition are powerful learning tools, and I've learned stuff by adhering to their philosophy. Plus they have some excellent and useful sentences, like Uxor maritum senilem habet ("The wife has an old husband") and Velisne vinum rubrum ("Would you like red wine?").

Overall, though, I'm not sure if the Duolingo/ALM approach can ultimately work all by itself. Maybe for others, but I guess I'm not going to give it a chance to be my sole way to learn Latin.

[1] By coincidence (or was it?) I started seeing sentences involving graves around Halloween last year.

[categories]   ,


  09:40 PM

As I said in part 1, when I started with Duolingo, I realized that it used the principles of the audio-lingual method (ALM), which emphasizes patterns and practice over grammar instruction. But how do you teach someone a language starting from nothing? And, as per the ALM/naturalistic philosophy, without explaining anything?

When I started with Duolingo, I didn't know how the app worked. Now I'm sorry that I didn't start capturing screenshots right at the beginning. But I'll try to recapture some of it by showing you roughly how the teaching progresses.

1. Start with pictures

Duolingo started by introducing basic vocabulary with pictures, like this:

As you can see, they make these as easy get as possible. Hey, Latin is fun!

2. Solicit sentences where it's hard to make errors

The next phase is to have you recognize words that you've been introduced to. They do this in a couple of ways. One approach is the classic multiple-choice answer. In my exercises, there have only been 3 choices, and I've found that the 2 incorrect answers are pretty easy to eliminate. They're not trying to trip you up; they're trying to make you successful, so they're asking for just a tiny bit of effort. So far.

(By the time I got around to snapping this screen, I'd gone quite a bit beyond the intro to the words puer ("boy") and urbs ("city"), obviously.)

Another approach that they use is to give you a short sentence and then have you assemble the translation by selecting (clicking) words from of a limited set of choices. In this phase, they again give you choices that make it pretty easy to get the right answer.

They use this approach a lot, and it's one way that they introduce changes in the pattern. Here's a variant form of urbs in a sentence:

In the first example, Corinna built four cities. In the second example, I built the city. What you're supposed to deduce from many repetitions of these types of related sentences is that when "city" is singular and the direct object, it's urbem; when it's plural and the direct object, it's urbes. But as I keep saying (sorry), they never utter terms like direct object, or for that matter, singular or plural. With enough repetition, you start picking up patterns like these.

They also use this "click the word" approach to introduce vocabulary that doesn't lend itself to pictures, like verbs. They give you a sentence where, by process of elimination, you figure out which one is the one you don't know. In this example, it's almost impossible (imo) not to figure out what venis means, given the choices they provide:

Up to now, you're just clicking. This seems weak—as someone on Twitter said, "I dislike clicking things to try to learn a language." A fair point, but there's a method here: you're simply seeing vocabulary, with a minor boost of having to actually pick from a small selection of choices. Now they change it up.

3. Solicit sentences aurally

The next step is that they dictate a sentence to you and you type it out:

This does a couple of things. One is that you encounter the new terms in a different medium, namely through the ear. (If I were studying a modern language, this listening skill would be critical[1].) An important benefit is that you're writing the terms that up to now you've only been reading and occasionally clicking. This is stepping up your language acquisition; you're now actually producing the language, albeit in a highly prompted way.

4. Translate to English free-form

Another form of exercise is where you produce a free-form English translation of what you're reading:

This takes away the training wheels (no hints via the clickable words) and it exercises your ability to understand what the different word forms actually mean. For example, in this example, you have to recognize that sunt means "are" and that Philadelphiae means "in Philadelphia," even though there's no "in" in front of it. (These are things that you would have been introduced to and drilled on before you see this exercise.)

5. Solicit translations into Latin

The most advanced exercise I've encountered so far is when they give you a sentence in English and ask you to translate it into Latin with no clues at all:

This requires everything you've learned: what words to use, how to inflect them, and what order to put them into (lexicon, morphology, syntax). With Latin, that's as much as you'd theoretically ever need to learn, but I guess I'll see down the line.

How to make this work

Although you do these different types and levels of exercises, the information is not presented in this strict sequence. Each lesson mixes up these different approaches in a set of 10 exercises.

The teaching also relies on these principles:

  • You get immediate feedback as to whether you got the exercise correct.
  • If you get an exercise wrong, it's repeated later in the lesson, though not immediately; they seem to follow a practice of intermittent reinforcement. As far as I can tell, you can't finish the lesson until you've gotten them all right.
  • You repeat and repeat and repeat. Even as you make progress, you do the same exercise again and again.
  • The lessons are short, no more than a few minutes. Their idea is that you can spend 5 to 10 minutes a day and make progress as long as you do it every single day.[2]

I'm probably overlooking some aspects of the pedagogy here, but I think that this is the gist of how they've designed the lessons for progressive learning: see, copy, listen, produce. As I say, I'm not sure I'll get to a phase where I have to respond to a question by writing a free-form answer, which would be to create novel sentences that have no direct pattern. (That would be hard to machine-grade, I think?)

I did make progress with this approach. However, as someone's who's studied other languages before, I found the no-explanation approach a little frustrating. More on that in the next installment.

[1] Language mastery consists of four skills, in order of difficulty: reading, listening, writing, speaking. Latin might be different from other Duolingo languages in that they have you read sentences, listen to them, and write them; but so far in my experience, they don't ask you to produce any spoken language. Then again … conversational Latin?

As an aside, they don't explain Latin pronunciation; they just show you words and then say them, and you sort out how the letters correspond to sounds. That's pretty easy in Latin, although I did eventually discover that I was learning "classical Latin" pronunciation as opposed to "ecclesiastical Latin," which sounds more like Italian.

[2] These principles are similar to the ones used by Kumon for teaching math and reading.

[categories]   ,


  10:19 PM

A couple of months ago, I started learning Latin by using Duolingo, a language-learning app for your phone or browser. I sort of knew about Duolingo because my kids had been using it, one to work on Spanish and the other to work on Japanese. They'd shared with me some of the amusingly strange sentences that Duolingo produces, like "A cat does not play piano" in Japanese.

When I learned that Duolingo had a course in beta for Latin, I thought I'd give it a shot. I'd somehow managed to never study Latin, a language that's always seemed not only inherently interesting but useful for understanding Spanish and, of course, for grokking English word origins.

We pause briefly here for my favorite scene from the movie "Life of Brian":

As soon as I started Duolingo, I recognized the pedagogic technique they were following. When I was learning German in high school, we knew this as the audio-lingual method[1]. To quote one source, this "foster[s] naturalistic language acquisition in a classroom setting." The idea is that you learn (internalize) patterns of the language. They don't explain any grammar. Instead, the instruction teaches you short snippets that it then changes in very controlled ways so that you can follow along. Here's an example:

Femina domi est. (The woman is at home)
Vir domi est. (The man is at home)
Puer domi est. (The boy is at home)
Puella domi est. (The girl is at home)

The hope is that you internalize xxx domi est for "xxx is at home." Once they've drilled you on this for a while, they change the pattern:

Feminae domi sunt. (The women are at home)
Viri domi sunt. (The men are at home)
Pueri domi sunt. (The boys are at home)
Puellae domi sunt. (The girls are at home)

Now they've introduced the concept that xxx domi sunt == "xxx are at home". You spend a fair amount of time drilling these variations until you seem like you've mastered both domi sunt and the various plural forms.

At no point do they stop and say that est and sunt are verbs, and est is used for 3rd person singular, or any of that stuff that you get in traditional language learning. The "naturalistic" approach of ALM is supposed to follow how children learn a language (I guess?), since we all learned our mother tongue without a single grammar lesson.

How can you use this to learn a language from scratch? I'll show you in the next installment.

[1] When I got to college and was studying German, one of our professors who specialized in pedagogy was from Germany and had a strong German accent. Ever since those days I only hear this as "ze odd-yo ling-val messod."

[categories]   ,

[2] |

  07:49 AM

I will not be attending the American Dialect Society's 2019 conference, which takes place next week in New York. This means I'll miss the highlight of the conference, which is the annual selection of the ADS Word of the Year. The nominating and voting meetings are raucous events that are hugely entertaining, even if you otherwise don't care one single fig for the rest of the ADS conference. (It's true; ask my wife.)

So this week for Friday words, instead of the normal, I'm going to list the words that I personally have been tracking for 2018.[1] Since anyone can nominate a word of the year, I also offer these as nominations in case anyone wants to present them at the live session.

To reiterate the pretty generous ADS criteria for nominations for WOTY:

  • doesn't have to be brand-new
  • needs to have shown a rise in popularity in 2018
  • can be a multiword phrase or compound

I'll mostly use the categories that ADS uses for their nominations, except that I have no particular hashtag to offer, and emoji? Not me. I have not actually picked winners here, for any category or overall. Feel free to vote for your favorite(s)!

Content warning: The list contains offensive terms and terms with possible political bias.

Political word of the year

There have been a lot of interesting words from politics this year. Here are the ones I found that I thought bubbled to the top of the 2018 heap.

chain migration
The process by which a legal resident of the US can sponsor the immigration of close family members. This is a Frank Luntz-worthy coinage that distills a complex issue into a sound bite to stir emotion.

shithole country
A term allegedly used by Trump to describe countries from which the US should not be accepting immigrants.

perjury trap
A question asked under oath whose purpose is to catch a witness in a lie.

crisis actor
An actor who portrays a victim of a crime or disaster. Used by conspiracy theorists to claim that certain disasters (such as school shootings) were staged.

deep state
According to right-leaning people, a cabal of bureaucrats working to undermine the current administration.

bottomless Pinocchios
A rating for political lies that have been repeated 20 or more times (multiple instances of 3- or 4-Pinocchio statements).

blue wave
The anticipated (and fulfilled) lopsided electoral victory of Democratic candidates during the 2018 midterm election.

Digital word of the year

I actually have 2 lists here. One is for digital terms likely to be used in ordinary conxtexts. The second is for terms that I see a lot at work—that is, emergent in the field of cloud computing.

Denying someone an outlet, whether on social media or in public forums, for espousing their views.

To aggregate a thread of Twitter posts into a single long-form page, using the Thread Reader App.

touchless technology
The use of gestures alone to control devices.

smart speaker
A wireless device that accepts voice commands and plays responses.

And here's my list of more "inside baseball" digital terms for 2018:

machine learning
A form of computing in which the computer system “learns” to perform tasks (such as identifying faces or predicting behavior) based not on prewritten algorithms, but instead based on analyzing a very large number of examples and deducing patterns.

An approach to software development that integrates development (programming) and IT operations in order to streamline delivery of features.

edge computing
A computer design in which the processing for the system is decentralized by distributing some of it to the “edge,” such as to IoT devices.

A form of cloud computing in which the mechanics of allocating compute resources (etc.) is left to the cloud provider, leaving the developer free to just write application code.

Slang/informal word of the year

My categorization starts to break down a bit here; the assignment of categories is a bit arbitrary.

To harass someone by calling in a false report of a crime at the victim’s address in the hopes of having emergency services respond (for example, a SWAT team).

Someone who is “involuntarily celibate”; generally associated with a subculture of men who hold (sometimes extremely) misogynistic views.

Most useful

Ineffective (“effect-less”). An old term that got a boost from Samantha Bee’s characterization of Ivanka Trump as a “feckless cunt.” (NB: the word “cunt” has a substantially milder connotation in the UK, especially Scotland, than it does in the US. For details, see the Strong Language blog.)

The cohort of people born between the Generation X and Millennial generations (late 1970s through early 1980s).

Most likely to succeed

birthday fundraising
The act of “donating” a birthday by asking well-wishers to support a charitable cause.

shadow banning
To restrict the visibility of a user’s social media posts without the user being aware of it, thus limiting the user’s reach without actually banning them.

Most creative

A person who is skeptical of the weight that was reported on the president’s medical report released in January. Based on the term birther for those who were skeptical about Obama’s birthplace.

“Confidently lecturing someone about a complicated issue on the basis of having watched one Ted talk about it.” (@JamieJBartlett on Twitter)

Euphemism of the year

executive time
Officially, unscheduled time on the president’s calendar, but widely thought to refer to the time the president spends watching TV or tweeting.

tender age shelter
The cages in which children are kept after being separated from their parents at the US border.

Individual 1
The unnamed owner of a company that the lawyer Michael Cohen worked for for 10 years, according to a court filing for the charges against Cohen.

WTF word of the year

emotional support peacock
Narrowly, a bird that a United Airlines passenger attempted to bring onto a flight as a therapy animal. More broadly, the point at which emotional support [creature] jumped the shark. (Ana Navarro in a tweet: “I think I need an emotional support peacock.”)

A viral sound test that asked listeners whether they heard “yanny” or “laurel.” (Jason Kehe of Wired used the term generically to describe the low audio quality of modern TV: “yanny/laurel times a million.”)

This is so sad Alexa play [artist/song]
A meme that represents ironic sadness, based originally on an innocuous tweet that someone posted when the cat ate their dinner.

bone saw
A normally unremarkable medical instrument that became a strange and unavoidable part of the narrative about the murder of the the Saudi journalist Jamal Khashoggi.

The full list

The lists above are culled from a fuller list of terms I was tracking. For completeness, I'll go ahead and list those here with minimal explanation, less the ones I've already listed.

[1] Other people have written about their personal WOTY nominations (for example, Nancy Friedman), but I've made it a point not to read those before I posted mine. I'm sure as soon as I see others' lists, I'll smack my forehead and exclaim "How could I have forgotten about that?!"

Like this? Read all the Friday words.

[categories]   ,


  10:50 PM

There was a flutter of discussion recently on social media (again) about who and whom. Mary Norris, an editor formerly of the The New Yorker, argues "So does civilization depend on the vulnerable 'whom'? Yes." John McIntyre, long-time editor at the Baltimore Sun, instead argues Just use "who"—in other words, forget about whom.

A not uncommon pro-whom position is that it's important to maintain the distinction for the sake of clarity. Is that true? I've been thinking about something I read recently about case marking in English, which is what the whole who/whom thing is theoretically about. We case-mark some words in English—that is, we change their form to indicate whether they are the subject or object in a sentence:

  • She called him and then sent them a text.

She is the subject; him and them are objects. We use specific forms of the pronouns here to indicate who's doing what. Just for fun, let's create a Very Incorrect Sentence:

  • *Her called he and then sent they a text.

I think we agree that in this sentence, the case marking for the pronouns is all wrong.

But let's look at a similar example:

  • Mary called John and then sent the editors a text.

Same sentence, only this time we use nouns instead of pronouns. Which is the interesting point: in all of English, we use case marking to distinguish subject and object for a mere seven words: I (me), he (him), she (her), we (us), they (them), who (whom), and whoever (whomever).

Notice what's missing in this list of case-marked words:

  • Other pronouns (you, it, the indefinite one). Unlike he, she, we, et al., these pronouns don't have distinct forms for subject and object.
  • Possessive pronouns (my, your, his, hers, its, our, their, mine, yours, ours, theirs, whose). Same: one form to mark them all.
  • Relative pronouns (that, which). Ditto.
  • Reflexive pronouns (myself, yourself, himself, herself, itself, ourselves, themselves, oneself)
  • Determiners (a, an, the, this, that, these, each, every, any, all, some, no)
  • Cardinal and ordinal numbers (zero, one, two, second, twelfth)
  • Adjectives
  • Nouns

Why is this significant? Well, English used to mark all of these types of words for subject and object, back when English was still Old English/Anglo-Saxon. And languages like German and Russian still do mark them (well, nouns less so) even yet today. But we've lost subject and object marking on all of these types of words, all except that short list of seven, and we seem to be able to understand sentences just fine.

Here are a few example in which words of these types are acting in different roles. I've marked all of the terms that are acting as subject or object.

  • The owner herself gave it a thorough cleaning with the new vacuum cleaner that she just bought herself.
  • After Mary called you, she texted you the address. Did you get it?
  • My new phone beats my old phone by a mile.

Again, except the couple of words from the short list of seven marked words, nothing is explicitly marked. To pick out a few instances, in the first sentence, herself refers once to the subject and another time functions as an indirect object. In the second sentence, you plays the part of (in order) direct object, indirect object, and subject. In the third sentence, my works for both the subject and the direct object.

Is there any ambiguity about what function any of these words play in the sentence? No. We don't need to mark any of these words as subjects or objects because we can tell from just the word order. Consider these sentences:

  • My new teacher called you.
  • You called my new teacher.

Identical forms of the words, but the order of the words tells us who the subject and object is. That's true for all of the example sentences. In fact, you could even go back to the Very Incorrect Sentence and argue that in spite of the pronouns being all wrong, it's still pretty clear who did what, because we can make a very good case (haha) based on just the word order.

What we're experiencing in English is that who and whoever are moving off the already short list of case-marked words. In conversational English, and as McIntyre suggests, we get by without whom. It's hard to argue that this is affecting our ability to understand the role that who plays in a sentence. Have a look at these sentences:

  • Who did she call just now?
  • Who did you give it to?
  • Who does the insurance cover?
  • Tell me about the person who you met today.
  • I have three brothers, one of who is a doctor.
  • We don't know who we will hire.
  • Who's fooling who?

All of these sentences use who where it "should" be whom. But it's not possible for a native speaker to misunderstand the intent behind each.

If you want, you can maintain that whom has a place in English grammar, and we should use it correctly in formal writing. If you do, though, be clear about why losing whom is an issue. Maybe civilization hinges on the use of whom, as Mary Norris posits. But it's not because we're going to be confused if it follows the path of so many English words before it.


[3] |

  10:20 PM

Over the Thanksgiving holiday I got another chance to eavesdrop at leisure on grandboy J’s language development. (I did part 1 in April when he turned 2.)

The tl;dr is that he’s progressing quickly, as one would expect from a kid who’s 2-1/2 years old. He talks up a storm, and he’s reached a point where you interact naturally with him using language—that is, you talk to him assuming that he’ll understand you, and virtually everything he says makes sense. Of course, he makes errors, but they're interesting because they seem to tell us something about language development.


J has trouble with unvoiced th (/θ/), which he often pronounces as an /f/ (“wif,” “I’m firsty”). I thought I detected that he can produce soft th sounds (/ð/), as in the and this. This would not entirely make sense, and it's possible that for /ð/ he's producing /v/ sounds and I just wasn't hearing it.

He also has issues with /r/, which he sometimes pronounces as /w/ (“weally” for really). I didn’t pay close enough attention to determine whether he always does this.

J has trouble with some other clusters as well. The one that struck me was “code” for cold. There must be something systematic about that one, because I’ve heard some adults do something similar, like “woof” for wolf.


Those are Triceratopses.
Generalization of -s/-es as plural.

Guys, look!
Imperative; second-person plural vocative (guys).

Opa, knock down!
Used to mean both “I am knocking you down!” and “Knock me down!”

Can you take[ ]apart it?
Not yet recognizing take apart as a phrasal verb, or just a mistake in moving the particle to the end.

I don't know what happened.
Do-negation (don’t), subordinate clause.

Can I have it back?
Yes/no question inversion, have back as phrasal verb

George wants to go on a walk by himself.
Auxiliary (wants) with infinitive (to go) that’s part of an idiom (go on a walk); adverbial prepositional phrase with reflexive (by himself).

[Adult]: Do you want some almond croissant?
[J]: I want so much.
[Adult]: How many cashews do you want?
[J]: I want so many.
Distinguishing mass and count nouns (so much/so many == “a lot”)

I ate it all gone.
all gone == all up, presumably a generalization of something like It's all gone.

I want to go see who is that.
Auxiliary (wants) with infinitive (to go see), subordinate clause. J used wh-question syntax for the subordinate clause where who is the predicate nominative—interesting error.

[Adult]: He’s going to eat you!
[J]: I don’t want to be eaten!
Passive transformation in a clause with an auxiliary verb. This one impressed me.

She did a good job giving my hair a cut.
Possibly a generalization of indirect object (give [indirect-object] a [direct-object], e.g. give me a toy). But correct use of a gerund phrase (giving) following "good job [of]."

Irregular verbs

I bit it
The dragon flew away
I ate it

Correct use of ablaut in irregular verbs. But …

I drawed it
I breaked it

Generalized -t/-d applied to irregular verbs. And …

We camed over here
Blend—ablaut and dental.


J is still learning the semantic space for different words. We mostly noticed this because he seems insistent on using (and having others use) specific terms.

[J]: This is my digger.
[Adult]: Is that a backhoe?
[J]: No, it’s a digger.

But the next day …

[J]: This is my digger.
[Adult]: Oh, that’s a front loader.
[J]: This is my front loader!

J is going to another room.
[Adult]: Bye-bye!
[J]: I'm not leaving!
Bye-bye is reserved for going home or an otherwise more permanent parting.

[Adult]: Can you help me undo the Velcro on your shoes?
[J]: Those are straps!

Polite speech

J wields a number of phrases and sentences that seem to derive from adults’ corrections for tone and politeness. He's in preschool, which is probably the source for some of these.

Can I borrow that real quick?
i.e. Can I have that?

No, thanks! No, thanks!
Asking someone to stop tickling him

How's your day going so far?

[categories]   ,


  07:48 AM

Here we are in August, which reminds me that the name of the month is a capitonym—a word that changes meaning depending on whether it’s capitalized: “The august professor was born in August.”

I have two new-to-me words this week that are related to shapes. The first is scutoid (apparently pronounced SCOO-toid), which is a remarkable thing: a heretofore unknown geometric shape. I mean, you’d think by now we’d pretty much found them all, right? The actual shape is a bit involved to describe, so I’ll lift the definition and more from the article where I learned about this: “prism-like, with six sides one end, five on the other, and a strange triangular face on one of the long edges of the prism.”

Something I found interesting was that scientists modeled geometries to determine which shape would fit together best when arranged both flat and in a curve. Then they went looking for that shape, and they found it! Apparently it’s all over the place in nature. Not only did they predict the shape and then find it, they got to name it. The name is based on the scutellum of a beetle, which is sort of the carapace of the insect.

A second shape name came to me recently via Friend Ralph on Twitter. He pointed me to a blog post that mentioned a lemniscate, which turns out to be a formal name for a figure-8 shape. And by formal, I mean there’s a mathematical description of how to create the shape, as determined by mathematicians starting in the 18th century. The name comes from Latin (of course), meaning in effect “beribboned”; the lemni- part derives ultimately from a word for ribbon, which is a nice visual for the lemniscate shape.

New technical words are maybe not all that interesting, but what struck me was that the blog author had used lemniscate metaphorically. He’d devised an idea that the lobes of a lemniscate represent quasi-opposing camps (in his case, progammers versus IT/ops people), at one point writing how developers “hopped to the other side of the philosophical lemniscate.” Here’s his representation:

I have some darkish thoughts about the use of an obscure term like lemniscate in a blog post, but I guess I should just be happy to have been introduced to this term, as metaphor and otherwise.

It's nice to sit around with friends and discuss things, right? Etymologically, maybe not so much. The word discuss has a more violent origin than you might think: the very original Latin meant "to shake apart" or "break into pieces." However, already in late Latin the word was used in legal contexts, where it referred to examinations and trials, and we got that sense from our friends and conquerors the Normans. It then evolved into the milder sense of "talk over" that we now have. Tho of course at times some "discussions" might indeed hearken back to the original sense.

Like this? Read all the Friday words.

[categories]   ,


  10:03 AM

Politics surely is a rich source of new terms, even if most of them are weasely. This week I saw an article about James Allsup, a prominent alt-right personalty. Allsup had been called a white supremacist, and various GOP officials in the state of Washington had officially distanced themselves from him. But in private, a local party chair who supported Allsup said that the candidate had been label-lynched.

There are a number of interesting things about this term. The connotation is that the Allsup had been metaphorically killed via language, moreover with the idea that this had been done extra-legally. Dictionaries I've looked at don't list metaphoric meanings of lynch, but it's not the first time that the word has been used like this; Clarence Thomas used the expression high-tech lynch mob to describe (and, as some say, shut down) uncomfortable questions that came up during his confirmation hearings for a seat on the Supreme Court.

The word lynch is a very fraught term in the US. It invokes senses of mob rule, of enduring and extreme prejudice, and of innocent victims. And of course lynching was prevalent for a long time in the US as a largely unpunished crime that was used to exert violent and unjust control over a minority population. Invoking the word lynch is serious business. So it's some kind of verbal jujitsu to use a term like lynch to describe the reaction to someone's white supremacist views. Not to mention that this is paired with label (label-lynching) to describe someone who routinely uses terms like cuckservative, along with an insulting set of terms to describe African-Americans, Jews, and women.

The term seems to be relatively new. An article suggests that it has currency in the alt-right community, and was possibly invented earlier this year. The Spokane newspaper that broke the story of the GOP chair admiring Allsup might be the term's entry into a wider world.

I will say that as a piece of language, the expression label-lynching feels like something that could have been invented by master propagandist Frank Luntz. The alliteration, the bumper-sticker mentality, the implicit outrage: these all feel like attributes that can give a term like this legs.

Ok, enough of that unpleasantness. Let's move to origins. A tweet this week by the folks at Dictionary.com clued me in to an unexpected etymology for the word penthouse. It was not originally a house and it wasn't, um, pent.

The word as we got it from French was apentiz, which referred to an attached building or lean-to. This is related to appendix in the sense of "attached." But two things happened. One was that the initial and unstressed a- dropped off, a process known as aphaeresis or aphesis (compare around > 'round and excuse > 'scuse). That left us with a word like pentyz (various spellings).

Then a process called folk etymology took hold. The word pente meant "slope," and people heard "pent-is" referring to a building with a sloped roof, and they thought that the -is part must actually be -house (hey, a building, right?). So the word actually turned into penthouse. It wasn’t till the 20th century that penthouse was applied to a (small) apartment or structure on the top a tall building. And from there the normal rules of real estate converted location, location, location to luxury.

Like this? Read all the Friday words.

[categories]   ,


  08:10 AM

Last week there was a technical conference about cloud technology that a lot of our colleagues went to. As they do, people live-tweeted about what they were seeing. At one point, our boss tweeted an observation about the term on-premises:

This was a wee bit of a joke. Those of us who work in cloud technologies talk about on-premises resources, which refers to stuff that isn’t in the cloud, i.e., that's on the customer's site. And we’ve been adamant that it’s is on-premises, with an s at the end, not on-premise. The word premise refers to a proposition or basis ("The premise of the TV show is that …"), which is quite a different meaning than premises, which refers to the space occupied by a business ("No drinking is allowed on the premises").

But in our editing we change on-premise to on-premises all the time. Which is to say, s-less on-premise to mean on-premises is widespread. This means that people don't really think about what the component pieces of on-premise(s) really mean; they're using on-premise as a single term. In language talk, the expression has been lexicalized with idiomatization: the expression has been taken into the lexicon as a unit. (Compare could care less, as in "I could care less.")

In the same spirit that Jim posted the tweet, I suggested that on-premise would be the beg the question of 10 years from now. By which I meant that on-premise would be so widespread that people didn't even realize that this was technically a mistake.

Of course, we could solve the problem at a blow by just going straight to on-prem ("Migrating from on-prem to the cloud"). I think of this as the alum solution—who can keep track of alumnus/alumna/alumni/alumnae? No one, that's who, so let's just go with "They're all alums of the University of English Spoken On-Premise." :)

Update There's an interesting discussion in a comment on Adam Fowler's blog about why s-less on-premise makes sense morphologically in English.

And another update! Katherine Barber, a Canadian lexicographer, addressed the premise/premises question some year ago. Her conclusion:

In fact, if you do a Google search on "licensed premise" you find the term in many legal documents, from all over the English-speaking world. If this usage bothers you, my advice is: hie you to a licensed premise, drink up, and accept the inevitable.


[1] |