A chunky good man

My first instinct when confronted with a hot political controversy is to go over the language used to express it with a fine-toothed comb*. It is in this spirit that I noticed the great frequency with which the chunk “a good man” being thrown about in reference to the recent Kavanaugh hearings, either for Kavanaugh himself or good men in the abstract.

Example courtesy of Lindsay Graham:

This good man should not be destroyed. If you legitimize this process by one vote short, woe be unto the next person.

My first thought was that if they had nominated a woman like Amy Coney Barrett instead, and she had had similar alleged incidents in her past, this rhetorical nugget would be unavailable for her defense. No phrase of similar cultural heft exists for women, although the phrase “a good woman” is just as grammatically possible as “a good man”. My guess as to why is that while the image that the phrase “good man” conjures up in people’s minds is an archetype of competence, dependability, and bonhomie (emphasis on the homme), “good woman” only vaguely summons the idea of something like a loyal wife. Woe be unto any woman nominated for a high position who needs her character defended with reference to implicit cultural norms.

blur close up focus gavel
PIctured: The tool of a good man, not a good woman. Photo by Pixabay on Pexels.com

I did a quick search of BYU’s corpora to see if the linguistic record backs up my instincts.

The string “a good man” gets 12372 hits on iWeb and 1643 on COCA.

Screen Shot 2018-10-05 at 10.00.14.png
I’ll leave it to you to dig into the contexts.

Meanwhile, “a good woman” gets 1807 on iWeb and 262 on COCA.

Screen Shot 2018-10-05 at 9.59.43.png

That’s almost 7x the frequency on iWeb (the larger of the two corpora) and more than 6x in the other.

COCA, unlike iWeb, allows you to separate hits by their source (magazines, academic journals, spoken, etc.), yielding a bit more information of what kinds of contexts “a good man” and “a good woman” are typically uttered in.

Screen Shot 2018-10-05 at 10.05.14.png

Screen Shot 2018-10-05 at 10.05.22.png

Within that about 6x overall on COCA, “a good man” is used about 10x as often in spoken contexts, 5x in fiction, 5x in magazines, 8x in newspapers, and almost 4x in academic writing. For some reason, Bush’s first term in particular also sees a spike in use of “a good man” – perhaps this relates to the politics of that time, including the 2004 election, where adherence to certain conceptions of manhood were a subtext for the Kerry and Bush campaigns.

For comparison, “man” and “men” occur a total of 582,307 times in COCA vs. 483,248 times for “woman” and “women”. This means that “a good man” does indeed occur much more often relative to “a good woman” than one would predict if the phrase were simply a matter of combining parts of speech according to the rules of grammar. “A good man” is a chunk bordering on an idiomatic expression for a certain, known, type of person, like a “people person” or a “person of faith”. This particular type of person’s goodness seems to depend on their not being a woman.

(* The iWeb Corpus lists “fine tooth comb” as about 3x as frequent as “fine toothed comb”, neither with a hyphen. This makes my usage rather pedantic to the point of being functionally incorrect.)

The simple present, unsimplified

Since I started my hobby/rigorous research pursuit of conducting Google Forms surveys on grammar, I have been thinking about the big one.  The one that combines the most assumptions and nuance and the simplest form into a wad of meaning with white dwarf-like density, which is maximally unbalanced in its complexity and the earliness and brevity with which it is treated in grammar textbooks.  The big one is, of course, the present simple.

This is going to be a long post.

Read More »

Grammar Mining (and the collected Mark SLA Lexicon)

Many of us agree that teaching “at the point of need” (as I believe Meddings and Thornbury put it) is an ideal context for formal grammar teaching.  Students’ trying to communicate something provides clear evidence that they need the grammar that would facilitate communicating it, and depending on how close they come to natural expression, evidence that their internal representation of English is capable of taking on this additional piece of information.

In interlanguage punting, I conjectured that taking a guess at grammar students may need in the future and organizing a lesson around a particular grammar point was justifiable if the lessons you used to introduce that grammar would be memorable long enough for a “point of need” to be found before the lesson was forgotten.  At the time, I was teaching weekly 1-hour grammar workshops with rotating groups students at different levels, and as I could not teach reactively I had to justify my grammar-first (formS-focused) approach.

Read on for the last post before the new semester starts.

Read More »

The Devil’s Dictionary of Correction Codes


Wrong in ways I can’t be bothered to specify.


Everyone from your junior high English teachers to your ESL instructor has tried to explain the differences between gerunds and infinitives to you using logic and rules of thumb.  We were just trying to make ourselves sound smart.


You accidently requested that the reader commit a human rights violation instead of informing them that one had happened.  I don’t have time or space to explain that, so here’s a single letter instead.


Please stop writing according what 99% of your input implies are the rules for native-like English.


The teacher is willing to treat this as a language error, but secretly believes you wouldn’t notice if you suddenly had 3 cats instead of 1.


Wait 100 years or so until the 3rd person singular dies out and this will no longer be a problem.


In English clauses, you don’t need to show degrees of formality, gender, intention, or whether the information in it was learned directly or indirectly.  However, you must always be clear when it happened (roughly divided into the past except when it’s relevant to the present, the present which isn’t really the present, and the future except in subordinate clauses) and remind the reader of that with each finite verb.  We’ll just assume you know what a finite verb is and which ones they are.


As an ESL student, you are expected understand and apply metalanguage that native speakers need to complete at least 2 years of post-graduate work in linguistics for.


I won’t let you transfer or get your second Master’s degree in engineering until you show respect for conventions of writing that are present in only 0.01% of natural input.


Just giving you the answer would save us both time, but making you do the work allows me to claim that my marks are student-centered.


Your teacher consulted Google and confirmed that this comma should not be here.  It might belong somewhere else.  Google it.


We expect your use of conjunctions to be more correct than the New York Times.


(I hope it’s clear that I’m making fun of teachers including myself here and not learners)

Dead quantitatives, revived by “grammar”

A bunch of teachers have taken my grammar test.

Or is it… a bunch of teachers has taken my grammar test?

Why does the second sound so bizarre?  Is the frequency with which we match apparent subjects like “a bunch” with “have” or “are” a lamentable pattern of grammatical laziness or is “bunch” just a special kind of word, rather than the noun it appears to be?

An interesting transition appears to have happened or be happening to English partitives and quantitatives, phrases like “a piece of”.  Under certain circumstances, they seem to lose their grammatical class as noun phrases and are instead interpreted like adjectives, modifying a noun to come rather than being nouns themselves.  You know the most common of these – “a lot of”, which appears to be a noun phrase with an indefinite article (“a”), a single noun (“lot”), and a prepositional phrase (“of ~”).  In practice, “a lot of trees” is interpreted as a noun phrase about “trees”, not about a “lot”, which can see reflected in in the verb conjugations in sentences like “a lot of trees are in the park”, wherein “are” conjugates to match the plural noun “trees”.  Needless to say, different noun phrases with a similar forms to “a lot of”, e.g. “a room with” or “a shot in”, are not treated this way – “a room with windows” is not a noun phrase about windows.  I’ve never heard a sentence constructed like “a room with windows were open to let in the breeze” – have you?

You probably think I’m about to condemn a lot of the English teachers who took my survey for having bad grammar.  No, I’m not.  Instead, I’m about to propose a semi-regular change in grammatical class that most people’s (including my) notion of correct grammar hasn’t caught up with yet.  I name this below the jump.

Read More »

Discursive stowaways and human logic (dangling participles part 2)

Dangling participles are less ambiguous than style manuals would have you believe.  They are subject to the same basic rule that governs all modifiers – namely, that human readers with functioning representations of the real world will give them the most plausible interpretations and move on.  At worst, they are just like a lot of adverbials or adjective clauses in that they could conceivably refer to multiple parts of the sentence.  More often, danging participles in common use are essentially idioms with set meanings, whether or not they share a subject with the main clause.  These are the ones you hear on the evening news – keep an ear out and you’ll catch quite a few.

I put together another survey after the last one to further investigate what may make a dangling participle seem more comprehensible or clear besides having the subject of the main clause as its subject.  Specifically, I was interested in a few things that seemed to be the most common implied subjects, and whether using these reliably made a dangling participle more comprehensible than other implied subjects.  My conclusion was not what I had expected.

If that hasn’t already put you to sleep, read on.

Read More »

The simple past in simplified history

I had an interesting conversation with a fellow dog-owner, who happened to be an Indian nationalist [Edit: Apparently the term for people of this persuasion is “Hindu nationalist”, not “Indian nationalist”.  Thanks Adi Rajan], at the dog park.  My interlocutor was recounting some of the wrongs that had been visited on Hindus in India by foreign conquerers, and he described how one named Aurangzeb had a particularly bad habit of tearing down Hindu places of worship and replacing them with mosques.  As it happened I had just finished reading Atrocities again and was sort of on the same page mentally, or at least more prepared than average to hear stories of Mughal emperors sweeping armies across the subcontinent, disrupting agriculture and failing to plan for floods, and generally causing a kind of misery that has political power hundreds of years into the future.  Oh, and don’t ask me how we got on the topic.

You might be wondering why the parasol-bearer is so badly failing at his job.  Actually, what he’s holding aloft is a massive lemon meringue pie, which Mughal emperors would order baked after a successful military campaign as a show of strength.

Anyway, he mentioned one countermeasure that Hindus took during Aurangzeb’s reign to at least be pillaged on their own terms.  As was explained to us, it was (is?) normal in Hinduism to cremate bodies soon after death, so that the soul didn’t have anything in this world to cling to when it has to move on.  In the case of holy men, upon (physical) death the bodies were kept and/or preserved rather than cremated.  This was, of course, because holy men’s souls can move independently of their bodies.  Holy men’s mummified corpses from that era would presumably still be on hand if observant Hindus hadn’t taken it upon themselves to cremate them as well during Aurangzeb’s reign, to prevent them from falling into the hands of the Muslim conquerers, in a bit of proactive self-desecration. This was, according to the man at the dog park, characteristic of Hindus, who always sought to keep their faith pure.

I got to thinking about how common this practice (let’s call it proactive saint cremation, or PSC) could really have been, as part of my usual ruminations on how in the creation of a group narrative, “a few people did it” turns into “people did it” and then “we did it collectively displaying the unique characteristics of our people”.

I realized that some semantic properties of the “simple past” (scare quotes for bad naming – it’s no more “simple” than the “simple present”) might enable this transition.  Namely, the blurriness of the simple past with respect to whether it refers to a single event or a stereotyped, repeated event facilitates the transition of historical occurrences from discrete to characteristic of a people, place, or time period.  The fact that the adverbials that serve distinguish the simple past for single occurrences from the simple past for repeated occurrences are easily discarded is of significance as well, as well as other qualifiers on the noun subject which are often grammatically inessential.

For example, let’s say this is a historically justifiable statement:

Ruling Muslims from the upper class ordered Hindu monuments destroyed in 1699.

(I’m not saying that this sentence is true – just using it as an example)

With the adverbial prepositional phrase removed, it is easily interpretable as referring to a repeated action.

Ruling Muslims from the upper class ordered Hindu monuments destroyed.

And with all the grammatically inessential (i.e., non-head) information removed from the subject noun phrase,

Muslims ordered Hindu monuments destroyed.

It would be plausible for someone just joining the conversation at this point to hear a blanket indictment of Muslims rather than a description of a particular historical event.

Now, part of what makes this possible is the particular grammatical feature of English that the same verb form, the badly-named simple past, works both as a past version of the simple present (i.e., it paints the subject with a stereotyped action occurring at no particular time, like “dogs bark”) and as a reference to a single action taking place at a specific time (which the simple present does as well, but less often – see “he shoots, he scores” or “I arrive at 6 PM”).  Of course, if you want to be very specific about the fact that an action was repeated, you could use alternatives like “Hindus used to burn their dead” or “Holy men would be preserved instead”, but the simple past in the absence of qualifying adverbials leaves either interpretation open, and therefore makes extension of historical events from single and limited to common and characteristic very tempting.

Also driving this, of course, is the omnipresent impulse to narrativize one’s national history and define one’s or someone else’s ethnic group with characteristics that are “proven” with reference to stories like the above.  In fact, my inkling is that any ambiguity in descriptions of historical events will always be used to simplify them for inclusion in one country or another’s national story.  In Japanese, it is the lack of plurals for nouns, allowing “a Japanese apologized to comfort women” to become “the Japanese apologized to comfort women” with no change in wording.  I assume other languages have similar ambiguities that can ease the transition from events that happened to national triumphs or tribal enmities.  Grammatical ambiguity as in the simple past may be but one of many forms of catalyst that make historical events into parts of a story about us.

Dangling participles, rescued?

I was doing some staving-off-the-decline-of-my-brain reading when I came across this striking sentence:

Inching closer to the classroom, ideas are needed as to how explicit and implicit learning can best be harnessed and brought to bear on the acquisition task in a fashion that allows efficient progress and does not entail attempts to combine explicit and implicit teaching as separate endeavors with conflicting theoretical underpinnings, as has sometimes been proposed (see, e.g., R. Ellis 1993; Fotos & Ellis 1991; Willis 1993). (pp. 52-53)

Do you see what I saw?  I don’t know why my radar was up in this regard, but that sentence contains a dangling participle.  That is, the implied subject of the phrase “inching closer to the classroom” does not appear in the nearest independent clause ideally in the subject position adjacent to the participial phrase, or anywhere else for that matter.

(side note:  I also mean someday to investigate usages of “but” meaning “please regard the previous clause as a hedge”, as in the first line above.)

The quote comes from none other than Mike Long’s Second Language Acquisition and Task-based Language Teaching.  I wanted to see if my participle radar was mistuned, so I copy-pasted the relevant part of that quote on Twitter and asked whether my followers (that sounds pretentious) wires were tripped in the same way that mine were.  I got into an interesting discussion, mostly with Tyson Seburn and Rob Sheppard, of the issues involved, and realized there was more to what made a dangling participle seem odd than a missing subject.

I had some hypotheses about the results.

Hypothesis #1: Dangling participles are less likely to trigger alarms when they are part of a well-worn and familiar phrase like “having said that”, regardless of the contents of the main clause.  Rephrasings of those idiomatic expressions will be judged as less acceptable even when their literal meanings are similar.  This is for well-known principles of idiomaticity, the same reasons we don’t demand normal grammaticality of phrases like “be that as it may”.

Hypothesis #2: Dangling participles seem less abrupt when their implied subjects are either the speaker OR the discourse/conversation/train of thought itself, even when these are nowhere to be found in the main clause.  This came from a bit of reflection that A) the speaker and the discourse are always potential, ambient topics – discursive stowaways if you will – and B) I know many transitional phrases and discourse markers in Japanese work like this (although of course nobody considers this ungrammatical because not only participial phrases but all kinds of phrases in Japanese often lack subjects).

Perhaps also lending support to hypothesis #2, many familiar, idiomatic participial phrases (hypothesis #1) have the speaker or the discourse as their implied subject.

Hypothesis #3: Barring either of those scenarios, people will scan a sentence for a possible subject and try to make the dangling participle make sense.  The danger in these cases comes when two plausible interpretations exist, as in “Running the light, Joe was hit by a drunk driver”, and not so much in Anguished English fodder like “Drooling on his shoes, Joe was greeted by his Great Dane”.

I base this hypothesis on the fact that we don’t make the demand of close proximity on all potentially ambiguous modifiers in English.  Take prepositional phrases, for example – we don’t require that “I painted a picture of her in the late afternoon” mean I painted a picture depicting the afternoon (whenever I put brush to canvas), while “I painted a picture in the late afternoon of her” means I had my brushes out between 4 and 7 PM.  We are usually capable of drawing out likely the most meanings through context, and I presume that when push comes to shove we’ll do the same for dangling participles.

Hypothesis #4: When the dangling participle is separated from its subject by an empty pronoun like “it” or “there”, people will disregard the empty pronoun and treat the next noun as the subject.  This is because people will regard the empty pronoun as simply a placeholder and skip over it in their search for a referent for the participial phrase.

Hypothesis #5: When the implied subject is also the implied agent in a main clause in the passive voice (e.g. Spraying water, the fire was extinguished), the dangling participle will seem less jarring.

As we shall see, not all of these were correct.  Or maybe they are, but I’d need a sample of more than 25 to say so.

I made a quick Google doc with a Likert-style 5-point scale of acceptability for 8 different sentences, all with dangling participles (still viewable here), tweeted the link around, and after a few days, compiled the results.  The survey started with these instructions:

This survey presents participial phrases that would traditionally be considered “dangling”; that is, the participial phrases do not refer to the nearest noun. Please indicate how correct or incorrect you would regard these utterances in *written contexts*. Note that this is not only a grammaticality test; all of the utterances contain something that is traditionally considered an error. In your assessments, you may include or exclude considerations of grammaticality, comprehensibility, conventionality, or other factors at your discretion.

As the comments indicated, not everyone read this part.  Anyway, I wanted to make clear that I was after more than just grammaticality.  That part at least seemed clear to respondents.

Here are the sentences I wrote:

Rolling down the street, I saw a car.  This is pretty close to a canonical dangling participle.  The subject of “rolling down the street” should be “car”, not the nearest noun “I”.  Still, the subject of the participial phrase is clearly present in the main clause, which may help its acceptability.

Standing at the podium, the students prepared for the teacher’s lecture.  Unlike the previous sentence, the subject of the participial phrase is only present in the main clause as a possessive, i.e. not as the head of a noun phrase.  If my hypothesis #3 is correct, the reduced salience of the subject sentence should be less acceptable than the last, although they are equally incorrect from a strictly grammatical point of view.

Pouring concrete, there is a construction crew in the street.  The subject is separated from the participial phrase by a dummy pronoun (or adverb, I suppose).  If my hypothesis #4 is correct, this sentence will be judged more acceptable than ones whose subjects are simply missing.

Moving on, the next topic is sales quotas.  This combines my hypotheses #1 and #2 in that “moving on” is a well-used phrase and has the speaker or the discourse as its subject.

Screen Shot 2017-05-22 at 22.21.51.png
From COCA.  Mostly spoken, but clearly common.

Knowing Bob, it’s his family that is distracting him from work.  Same as the last sentence, “knowing (person)” is a well-worn phrase that has the speaker as its subject.  Maybe this item was redundant.  If so, the scores for one this and the last one should be about the same.

Looking everywhere, the key was finally found.  The subject is the implied agent of a main clause in the passive voice, as in hypothesis #5.  This should make it less objectionable to readers than a sentence where the subject is missing entirely.

Gradually culminating in true agriculture, plants were gathered from smaller and more prescribed areas around human settlements.
This one is tricky.  The subject is the process described in the main clause, although that process is never nominalized.  Further, it is wordy, which may work for or against it: for it in that people may forget exactly how the participial phrase was worded by the end of the sentence and settle for getting the gist, and against it in that it has no chance of ringing familiar like “having said that” or “moving on”.

Switching rapidly between topics, grammar was discussed in turn with semantics and pedagogy.  This is a wordy phrase and a wordy sentence, giving it low familiarity, but the implied subject is either the speaker or the discourse itself, which should work in its favor.

You know what I realized now that I didn’t include at all?  A sentence where the subject is not even hinted at as an agent or buried in a subordinate clause!  Something like “Swinging with great force, the ball flew into the outfield”.  Next time!

…Also, none of these participial phrases are past participles.  I wanted to keep it simple this time.


ss all.png

This is for all surveyees.  5 is “Totally acceptable”, and 1 is “Totally unacceptable”.  As you can see, there are two early leaders.

I asked surveyees for their relevant grammar experience and education level, but none of these produced significant changes in responses, so we can forget about them.  For no reason at all though, I’ll just say that my Twitter feed appears to be highly educated.

On the other hand, there was a small difference in responses for people who said “comprehension” was a criterion for judgment.  These were 18 of the 25 total repondents.

ss comp.png

As opposed to clarity, 9 of the 25 respondents (some people gave multiple concerns).ss clar.png

I never gave or asked for operationalized definitions of “comprehensibility” or “clarity”.  I assume comprehensibility meant that at least one plausible interpretation was possible, and that clarity meant that one interpretation was “clearly” more attractive than others.  When respondents gave a longer explanation that seemed to coalesce around one of these definitions, I treated them simply as either “comprehensibility” or “clarity”.  For example, “When it is understandable”, was reduced to “comprehensibility”.  As you can see, the “clarity” respondents rated all the sentences lower overall.  Since all the sentences were written specifically to buck the rules of clear writing, this is not surprising.

Only two respondents included grammaticality as criteria, so I haven’t bothered to compute averages for them.

So, the clear winner in producing acceptable dangling participles is idiomaticity or familiarity, particularly among people who valued comprehensibility.   Let me go through the sentences for detailed results.

Rolling down the street, I saw a car. 1.92 acceptability with all respondents (1.94 just for comprehensibility and 1.78 for clarity).  This was the second-lowest score.  Its acceptability correlated with the next sentence’s acceptability with a score of 0.71, meaning that people who accepted one usually accepted the other and people who rejected one tended to reject the other too.

Standing at the podium, the students prepared for the teacher’s lecture. 2.12 acceptability (2.22/1.67).  I wish I could say that this score and the last score support one of my hypotheses but since I neglected to include a “pure” dangling participle, it’s hard to tell whether including the subject in the main clause, but not adjacent to the participial phrase, caused these sentences to be evaluated more positively.  What we can probably say that including the subject in the main clause somewhere other than adjacent to the participial phrase doesn’t help its acceptability by much.

Pouring concrete, there is a construction crew in the street. 1.88 acceptability (1.61/1.11).  This surprised me.  The subject of “pouring concrete” is only separated from the participial phrase by a dummy pronoun/adverb, yet this sentence was evaluated the least favorably of all.  Perhaps the simplicity of the sentence overall invited readers to mentally compare it to an easily generated improved version (“Pouring concrete, a construction crew is in the street”), and therefore reject it due to its clear inferiority to this salient alternative?  Or is it too jarring to try to imagine a “there” pouring concrete?

Mysteriously, this sentence’s score was correlated with the score for the “gradually culminated” sentence at 0.51.  I have no idea what could be behind this.

Moving on, the next topic is sales quotas. 4.00 acceptability (4.22/3.88).  Yes, the popular idiom which refers to the speaker got high marks. 

Knowing Bob, it’s his family that is distracting him from work. 4.04 acceptability (4.22/3.78)  And the other one did too.  They were also correlated with each other at 0.59.

Looking everywhere, the key was finally found. 2.16 acceptability (2.22/1.89). Much like having its subject located elsewhere in the sentence as opposed to completely missing (see the first two items), having its subject implied as the agent of a passive construction doesn’t appear to help dangling participles become much less objectionable.

Gradually culminating in true agriculture, plants were gathered from smaller and more prescribed areas around human settlements. 2.88 acceptability (2.83/2.78).  I wonder how the scores would have been for this one if I hadn’t told respondents in advance that it contained a dangling participle.  Having considered that possibility, it’s also potentially true that the acceptability of dangling participles increases the longer the sentence overall gets, and the reader can no longer be bothered to look for grammar mistakes for a participial phrase that occurred 30 words ago – like in this sentence.

Thinking about this sentence has led me to hypothesize about a third discursive stowaway in addition to the speaker and the discourse itself, which is the content of another clause.  This is also what is referred to by “which” in sentences like “It rained all summer, which helped the grass even if it killed my tomatoes”.

Switching rapidly between topics, grammar was discussed in turn with semantics and pedagogy. 3.00 acceptability (3.22/2.89)  So a candidate for a 2nd-place-finisher after idiomaticity or familiarity is reference to the discourse itself as a potential saving grace of a dangling participle.  Regrettably, I don’t have a survey item that tests reference to the speaker without using a common phrase, so I haven’t really tested hypothesis #2 properly. 

To recap, here are my 5 hypotheses on how dangling participles can be made acceptable and how they fared:

1) Familiarity/Idiomaticity.  This one seems to have the strongest support.

2) Reference to discursive stowaways (where the implied subject is the speaker, the discourse itself, or *update* the content of another clause).  Possibly.  Needs more research.

3) Presence of the subject somewhere else besides adjacent to the participial phrase.  Probably not, or not by much.

4) Dummy pronoun/adverb.  Probably not.  Needs more research (see below).

5) Implied as the agent of a passive construction. Maybe a little bit.

From Netflix.

Conclusions and next steps

So if there were a tl;dr version of this post, it would be “dangling participles are only barely acceptable when they refer to a few consistently available referents, such as the speaker, the discourse, or the contents of the next clause.  They are much more acceptable when they are common idioms or set phrases used consistently the same way.  Otherwise, avoid them.”

I drew up this survey and compiled its results in less than a week, and some issues slipped through the cracks.  Here are some things I would do if I were to try this again:

  • Test for acceptability of dummy “it” defined later in the sentence (cataphoric reference) – e.g. “Panting, it was the dog who seemed to dislike the heat the most”.
  • Test for missing subjects who are not even an implied part of the main clause, e.g. “Moving in, the truck was full of furniture”.  If all of my hypotheses are at least a bit correct, this should have the lowest score.
  • Test more for the missing subjects being the content of the main clause, as in “Making a clear statement of its priorities, there was no comment from the White House on the ongoing crisis”.  This was my third discursive stowaway.
  • Test past participles, a la “Decimated, the Wu army’s families anxiously awaited their return home”.

Having posted this with mistakes, it has now been edited for clarity.

Native speaker errors and the origins of which they come from of

I’ve seen more writing by native speaker (or non-non-native speaker) students than non-native speakers recently, for the first time in maybe 12 years.  I have to say, native speaker errors are to non-native speaker errors what economic depressions are to lost wallets.  The roots of the issues stretch downward through Bloom’s taxonomy until you’re unsure whether you should start your critique with “this subject and this verb don’t match” or “things that are not can’t be“.

Here are some examples, paraphrased and otherwise altered for the students’ sakes of course.

“The numbers are frightening how many guns there are.”

Okay, so some of the legitimate uses of it’s to start a sentence imply that it’s okay not to have a clear referent for every pronoun (there’s another example).  Some students take the common injunction not to let pronouns slide by without definition as an injunction to replace every pronoun with a noun phrase, as in the above sentence.  The problem is, it’s frightening how many guns there are is a passable use of… dang, I forgot the technical term for this, but it’s when you cataphorically define the pronoun subject it later in the same sentence, like it’s fun to travel.  I’m not really sure why this usage of it is is kosher in formal writing, but clearly many students think it isn’t and are willing to sacrifice logic to avoid using it.

“When I make friends it helps me to expand my circle of friends.”

For some reason people seem to feel like an adverbial is a more elegant way of introducing a subject than simply naming it.   I see a lot of sentences that start with a subordinate clause, and then have it as the subject of the main clause where it means “the content of the subordinate clause”.  Many of these also feature circularity as in the above example.  The problem for me is how to explain that the adverbial really should just be the subject (“making friends helps…”) without sounding pedantic or condescending.

Lists composed of nouns, have unlike parts, and last contained another unlike part.

Of all the errors I see in native speaker writing, this is the most familiar to me.  My Japanese students were similarly confused about the usual symmetry required when splitting a syntactic tree with a conjunction like and.  The thing is, I think it takes a brain unusually attuned to structure to be able to understand an explanation of this in explicit terms – most people probably get it just from reading a lot.  I’ve had limited success diagramming sentences like “I bought him a coat and umbrella” or “I bought him a coat and her a scarf” to show why “I bought him a coat and a hat for myself” is ungrammatical (or at least a confusing garden path).


If I could choose a single lexical item that portends a badly written essay, it’d be effective, as in “the author effectively establishes mood throughout the story”.  Many students interpret the word how in their teachers’ “describe how author X does Y in story Z” like an Olympic skating judge and simply rate it on a scale of “badly” to “very effectively”.  How does Charles Dickens use dialect to illustrate his characters’ social class?  Effectively!  A gold star for you, Dickens!  This relates to the problem of meta-theses outlined below in that it actually says nothing but announces that something will be said.  I’ve tried to avoid using “effective” in my criticism as a result of noticing this, as well as its polar opposite, awkward, the go-to criticism for proofreaders unable to describe grammar except to other people as smart as they are.

Dangling modifiers

I would have been extremely happy if my students in Japan had used any adverbials at all besides “when S V”.  To the contrary, native-speaking American kids are so loosey-goosey with modifiers that often the interval between periods seems to have nothing but modifiers treading water, having given up on the dream of firmament.  This being a problem, since grammar to be corrected.

Comma splices are everywhere, I don’t always even bother to correct them.

I’ve even started noticing myself making them sometimes, they have started slipping under the radar.


I think I did this back in junior high – used the space normally reserved for a thesis to announce that I had a thesis.  This can take several forms:

  1. “The thesis of this paper is…” Not terrible if what follows is still a thesis.
  2. “This paper will cover these topics…” The supporting topics are left standing like Greek ruins consisting of columns with no pediment.
  3. “How does Dickens use dialect to show social class?” Good question, when your teacher asked you to write about it.
  4. “Dickens uses dialect to show social class very effectively.”


Machine translation requires human-like AI

Some of you alive in the 90s might remember an episode of Star Trek: TNG that is held as an example of the philosophy of language showing up in popular culture.  In this episode, the voyagers of the Starship Enterprise arrive on a planet where the inhabitants speak a highly allegorical language, using phrases about mythic or historical figures a la”Shaka, when the walls fell” to convey messages such as “oops” or “I see your point”.  As a result of these literal translations, the Enterprise’s crew members are forced to decipher what the dense metaphors mean contextually rather than in their normal English idiom as the universal translators usually supply.  Universal translators, as you can probably guess, are supposed to work with any language on the first encounter with that language or even with the species using it, and as far as I know this is the only episode where this particulary difficulty arises.


The problem is, if a universal translator can’t work with the very (infeasibly, as the article above points out) allegorical language spoken in that episode, it shouldn’t work with any language.  Even very closely related human languages use vastly different grammar and vocabulary to express greetings, thanks, obligation, and anything else under the Sol System’s sun.  To know that the verb phrase “thank you” is a show of gratitute in English (not a command, as verb phrases in isolation generally are), while an adverbial like “doumo” serves that purpose in Japanese, a universal translator would need to be a mind-reader before it was a translator, as there is no way to ferret out the fact that “doumo” and “thank you serve the same purpose from first principles or even from the grammar of that language (which universal translators don’t always have access to; they work on every language even on the first try). Moreover, it would need to do this mind-reading on species whose physiology it has never encountered before, meaning it would need to determine where the locus of that species’ cognition is, make intelligent predictions about how the patterns of (presumably) chemical synapse firings correlate to intentions, and map those intentions onto speech acts as they occur in real time.  The prerequisite technology for a universal translator is much larger than mere substitution and reordering of words, and approaches impossible, even by sci-fi standards.

In our world, people often discuss non-sci-fi machine translation like Google Translate as if it also were a scaling problem of existing technology, as if adding more of the same gears and cogs we already have would result in perfect language-to-language recoding.  In essence, people think the incremental improvement of current machine translation technology can save us from the years-long process of mastering new languages ourselves.  This post, with its oddly long prologue, is meant to argue that perfect machine translation would require a project of enormously grander scale than the visible inputs and outputs of textual language, and like the universal translators in Star Trek, would have a project of imposing complexity as a prerequisite, one whose implications would go far beyond mere translation.  In the case of machine translation that prerequisitive is a complete human-like artificial intelligence.

Read More »