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.
‘Twould be a gift to be simple
As the cliché goes, the present simple is neither present nor simple when applied to actions (rather than states, which are present if not simple). Let me briefly sum up some of the arguments why this is so, most of which have been said by smarter people elsewhere:
1. It is not present. Recall the kind of picture you have seen in an introductory English textbook, for example of a boy riding a bicycle, captioned “John rides a bicycle”. Now imagine that one of your students comes across a scene in real life that looks exactly like that picture one day, and inspired to use the English he just learned, turns to a friend and utters, “Look at that boy – he rides a bicycle”. His friend agrees and recasts – “right, he’s riding a bicycle. He must be on his way to school”. The student has trouble figuring out why, with the very scene the present simple phrase appeared under in front of him, he cannot use it and sound natural.
Ironically, the first choice for describing an action that you see presently and simply right in front of you is not the present simple. In fact, the present simple is uniquely unsuited for this – it describes an action taken yesterday, the day before and maybe tomorrow, but definitely not at this very moment – you need the present continuous for that. This leapfrogging of the current moment by the present simple creates a catch-22 in which any attempt by teachers or textbooks to visually describe a habitual action immediately puts in learners’ minds a scenario which would be better described with another tense or aspect. That is, when you see a car driving, the caption should never read “Uncle Tom drives” – it should say that he is driving, since the car in the picture isn’t just driven often, it’s clearly in the middle of being driven. It’s actually rather hard to depict an action that isn’t taking place at the moment the picture is drawn. Enjoinders that students should use the present simple to describe “something that happens again and again in the present” are confusing, because the present is precisely the time that the event you describe isn’t happening. Descriptions of the present simple often have to choose between being understandable or being accurate.
A notable part of the difficulty is that the present simple also arrogates some of the future. The timelines with Xs at set intervals that teachers often use to explain the present simple inevitably extend past now into the unknown. This is why I claimed before that the present simple applied to actions can never be 100% accurate; it almost always involves at least some unjustified confidence that what happened before will happen again in the future. All of the factual basis for a claim in the present simple can only lie in the past (again, where actions but not states are concerned). Still the listener hears that claim as a prediction of the future as well – if I say “Nick picks his nose” based on having seen him do it a few times before, I use the present simple instead of the past simple or present perfect to imply that he will do it again. The present simple smuggles a sometimes unfair and always unprovable prediction of future behavior from past behavior into a deceptively basic-looking sentence.
2. It stereotypes. Part of the reason the claims to the future that the present simple makes seem so natural is that the present simple makes the behavior almost a part of the subject’s identity. When members of my generation said someone “listens to They Might Be Giants”, it coated them in a kind of faintly nerdy paint, much moreso than if we had said that he “is listening to They Might Be Giants” or “has listened to They Might Be Giants” (it was usually a he). The most accurate treatment I’ve seen of the present simple in a grammar-oriented textbook is Side By Side (I think it was book 2), which features characters of different nationalities, drawn in loose approximation of their national garb, next to sentences declaring something like “This is José. He is from Mexico. Every day, he eats Mexican food, he listens to Mexican music, and he watches Mexican TV shows.” In essence, it used national stereotypes to scaffold the idea that although the action is not happening now, it is nonetheless part of that character’s pattern of behavior. In this sense the verb is really functioning more like an adjective, applying now and forever but also caught out of time. The simplest usages of the present simple to describe actions are always stereotypes of a sort (whether national or otherwise); it is only natural that one uses well-known stereotypes to teach it. One may feel that in some cases these stereotypes are proper or even prudent (it may help one understand the world to first know that Mexican people generally eat Mexican food and grasp the nuances and exceptions later), but it’s hard to properly call them present or simple.
3. It is not simple. The joke linked above, while not about the present simple per se, could have easily used it – and to describe a pattern of behavior consisting of a single (extremely noteworthy) act. Contrarily, another act might seem ill-suited for the present simple if it is performed much less than normal: If I sewed myself into a new onesie once per season, it would seem odd to say that I “change my clothes”, but if I changed my car’s oil with the same frequency, then a sentence with identical grammar could be used unmarkedly. On the other hand, it would also seem strange to say “I change my oil” if I kept the same 4-oil-changes-per-year frequency but just did them at random intervals throughout the year rather than as scheduled, or all in quick succession on December 31st. As it turns out, there are a lot of variables, even restricting ourselves to temporal considerations, that make the present simple seem more appropriate or less appropriate.
The main purpose of this survey was to get a bit of a grasp on what kind of pattern of behavior most reliably gives people the feeling that the present simple is suitable to describe the relationship between that person and that behavior. In particular, I mean to take a deep dive here into what really goes into the “habitual or repeated action” in so many descriptions of this grammar point. How often is “repeated”? How intentional does an act have to be to be “habitual”? How much does it matter, as poor McGregor might wonder, whether the act is memorable or thought to say something particular about its performer?
With those questions in mind, about a year ago, I started thinking up some situations that might reveal interesting facts about the criteria that people use in determining the suitability of the present simple to describe them. Below is a list of those criteria.
Objective frequency: frequency in simple mathematical terms, instances over time. For the purposes of the survey, I considered objective frequencies of more than 1 per year to be “objectively frequent”.
Relative frequency: frequency as compared to a baseline, provided either by culture or particular to that circumstance. This is a bit harder to judge – clearly, changing your car’s oil every day would be “relatively frequent”, but what about once a month, or once every 6 months? I tried to design scenarios that would fall fairly clearly into either “relatively frequent”, “relatively infrequent”, or average.
Regularity: having a fairly constant gap between occurrences. I made it explicit in my contrived scenarios whether the action in question happened sporadically or with consistent frequency. The sentence “he eats hot dogs” might apply differently to Joey Chestnut and a typical Wienerschnitzel customer.
Markedness: the extent to which an activity is memorable for a variety of reasons, including strangeness. You may call this the “you fuck one sheep” criterion.
Stereotyping: especially when paired with a category rather than a particular instance, the extent to which an activity can be said to characterize that category’s normal behavior. “José plays Mariachi music” may ring more accurate than “José plays polka music”.
Variation: independent of frequency, whether instances of the activity were interposed with a similar, but different, activity. People may judge “Bob eats pasta” less true if Bob only eats pasta at the end of the month when he runs out of his staple, instant ramen.
Intent: whether the act was done with the will of the performer or not. Some cases, like “leaves change color” or “Timmy plays video games” are clearer, but as we will see below, others are a bit ambiguous.
Futurity: whether the pattern is described as continuing after the present. Obviously, “Ken lived” has a strong connotation of Ken’s no longer being alive. Meanwhile “Ken has lived” seems neutral, and “Ken was described yesterday by his doctor as a marvel of human fitness” has the opposite connotation. Some of the scenarios below were described as clearly continuing into the future, while others were left ambiguous.
These criteria were first entered into a spreadsheets with values of 1 (true) or 0 (false). I later changed the value to 0.5 for one and -1 for others for reasons I will explain below.
The sentences themselves
It turned out to be rather tricky to contrive scenarios that had randomness in the above-described criteria. For example, a very marked behavior is almost always relatively frequent as well (I hate to keep bringing up McGregor, but in his case the one salient act was also clearly more than average). In addition to hitting a decent mix of factors, I also tried to avoid using the present simple itself in describing scenarios that I was then going to invite surveyees to judge as candidates for the present simple – that would be like asking people the likelihood that an action described with the past simple occurred in the past. Even sentences meant to project the behavior into the future had to be couched in justifiable statements of fact – “he will” had to be changed to “he has decided to” in order to stick to facts rather than conjecture.
Because the usage of the present simple I was interested in was about repeated actions, I avoided stative verbs like “have” or “like”, which are not habitual or repeated by any standard. A seeming exception is the “have” in “have seizures”, which is not stative (in the same way that “have a baby” is not stative – if you’re “having a seizure” or “having a baby”, you’re in the middle of a discrete action. Oh, and call an ambulance). I’m also not dealing with other, different usages of the simple present – e.g. describing set plans (“I arrive at 6”), quoting (“Pinker states that…”), idiomatic phrases (“Here she comes”), or as mentioned earlier, narration (“Pinker passes… shoots… scores!”).
(Actually, “like” is becoming an interesting word to explain telicity in the same way that “chicken” is to explain countability (chicken comes in a bucket, while chickens tend to avoid them). Young people now “like” many things in a day, especially if they “like” their phones more than their teachers.)
The survey started with these instructions:
This survey will ask you to agree with whether a given sentence is accurate in light of other information. Please answer whether the sentence is accurate as presented, not whether it would be accurate with more information or with modifications.
I showed surveyees the scenarios written entirely in the past simple or the present perfect followed by “Based on the above, it is accurate to say ” and short sentence in the present simple in quotes. Under that sentence was a Likert-like scale ranging from 1 (Strongly disagree) to 7 (Strongly agree). At the end of the survey were two open-ended questions,
Please explain some of the criteria that you used for determining the accuracy of the present simple to describe the situations in this survey.
Please describe your personal and/or professional involvement with English. (years studied, jobs, first language, etc.)
Here are the scenarios and the present simple sentences I asked surveyees to judge the accuracy of, followed by the values for those scenarios I entered for the criteria listed in the last section.
Bob, a middle-aged man living in Nevada, has been to Las Vegas on vacation every year except 2010 and 2014. In 2010 and 2014, he went to the Grand Canyon. He has made reservations to go to the Grand Canyon again in 2018.
“Bob goes to the Grand Canyon.”
Objective frequency=0, (less than once a year)
Regularity=1, (exactly every 4 years)
Variation=1, (interposed with trips to a different place)
Futurity=1 (he is described as going again this year)
Cromagnastan, a country, invaded another country in the 2nd century BC, as well as the same country in the first century, the 7th century, and 15th century CE. It also invaded a third country in 1625, a fourth country in 1765, and a fifth country in 1910. Between these times, it had no wars.
“Cromagnastan invades other countries.”
Relative frequency=1, (this was hard to determine, but I think 6 invasions in the lifetime of a country is more than average)
Stereotyping=0, (the country was presented as fictional, but it is based on China, which does have a stereotype of invading its neighbors fairly often. I didn’t think the stereotype would survive changing the name.)
Intent=1, (to the extent a country can carry intent from century to century)
Penny, an elderly woman, killed her own pet when she was a child. When she was in college, she killed her roommate’s pet. After reaching middle age, she killed three neighborhood pets and then no more until the present. Between the killings, she has had other pets that she did not harm.
“Penny kills pets.”
Relative frequency=1, (most people never kill a pet)
Variation=1, (the interposing act being having pets and not killing them)
Lauren, a 24-year-old woman, had three seizures when she was a teenager. She hasn’t had a seizure since then, but the doctor warned her that she would likely have one again in the future.
“Lauren has seizures.”
John is an elderly man in a vegetative state. John’s caretakers started taking him to the park last month. Last week, he went to the park 3 times on random days of the week with his caretakers. This week, they took him twice already and they have thought about taking him again one more time, either on a weekday or the weekend. The park is the only place outside of the care facility that the caretakers take him.
“John goes to the park.”
Stereotyping=1, (old men at parks)
Intent=0, (more than a few respondents remarked that this was salient)
Jim, a young boy, has been forced by his parents to brush his teeth twice a day for as long as he can remember. He brushed his teeth twice yesterday and once this morning, and his parents have not considered letting him stop.
“Jim brushes his teeth”
Relative frequency=1, (this was a judgment call – I think twice a day is a lot for a young boy given that he doesn’t want to do it)
Intent=0.5, (although he is being coerced, I don’t think his scenario is analogous in intent to Lauren’s seizures)
Jenny, a college student, has showered once a month for many years. She showered once last month. She showered earlier this month and has chosen the day that she will take a shower next month. On most days between showers, she has only cleaned herself with a dry organic scrubbing powder.
Relative frequency=-1, (in this case, because the frequency was not just unremarkable but remarkably low, I used -1 for this value.)
Regularity=1, (exactly once a month)
Markedness=-1, (same reason as above)
Howard, a middle-aged man, has left rat poison in his garden for many years. Since he started leaving the poison, each year during the summer, a raccoon has eaten the poison and died. Many other times, rats have eaten the poison and died.
“Howard kills raccoons.”
Fluffy is a feral cat in a city. He saw robberies 6 times in the last 3 years at random times of the year in the alley where he lives. He has seen 2 robberies so far this year. Other times, he has just seen people walking through.
“Fluffy witnesses robberies.” (this one caused a few comments – why use the verb “witness” when I only used the word “see” in the scenario? The answer is that the present simple “see” can be stative, and I didn’t want any confusion whether or not I was talking about discrete acts. Still, the substitution probably reduced the value of this question.)
Kelly, a high school student, has visited her great-uncle in a neighboring city on Thanksgiving every year for the past 5 years, and has told him that she will visit again this year. She has not visited any other members of her extended family.
“Kelly visits her great-uncle”
Over the 5 days or so that the survey was open to Twitter and LinkedIn followers, it amassed 42 responses, only one of which had to be struck because the user “just wanted to see the answers” and put “Strongly Disagree” for everything. As has happened before, some surveyees seemed to think I was testing their grammar and wanted to see what the “correct” answers were, but in most cases, even these participants logged responses of their own rather than just clicking randomly through. I suppose they were disappointed no answers were forthcoming (until now).
In a nutshell, objective frequency, stereotyping, and a lack of interposing similar actions were the most predictive of surveyees’ feelings that the present simple was accurate for the given scenarios.
You can check out the raw numbers here.
Here are the mean responses for all questions with their standard deviations.
There are 2 clear winners and 3 clear losers. As you can see, the losers don’t lose as much as the winners win: The lowest mean was the “seizures” question with 3.46, or just below the halfway point. This is expected given that I didn’t give any sentences that clearly contradicted the scenarios accompanying them, like “Tom played golf every weekend until he died last fall./Tom plays golf.”
I computed the correlations between each participants’ answers and a chart containing the values (0, 1, 0.5, or -1) for the criteria listed above. That chart looked like this:
I ended up with a bunch of correlations for each participant – one for each criterion in the chart above. This told me how the answers for a particular scenario/sentence combination compared with the criteria (objective frequency, stereotyping, etc.) for that scenario/sentence. For example, one participant’s correlations were 0.56 for objective frequency, 0.21 for relative frequency, 0.09 for regularity, 0.19 for markedness, 0.04 for stereotyping, 0.11 for variation, -0.32 for intent, and -0.44 for futurity. That means that objectively frequency actions tended to have their accompanying present simple sentences judged to be accurate, while scenarios that clearly stated that the behavior would continue into the future were likely to have the opposite effect. This gave me a sense of what criteria were important for each participant.
I then averaged these out to get mean correlations for each criterion. The results are in the chart below.
As you can see, the strongest correlations overall were for objective frequency, stereotyping, and variation. Variation is negative because the presence of a competing activity to the activity in question reduces the likelihood that people judge the accompanying present simple sentence to be true. This was as expected. I will get into more detail on the three strongest correlations below before rambling on some more about this and that.
Discussion and a few more results
The scenarios that I considered “objectively frequent”, with a frequency of at least once a year, were the last 6 items: park, teeth, showers, raccoons, robberies, and uncle (sorry for the shorthand – I don’t think we want to see the complete scenarios repeated every time I mention them). The mean response for only these 6 was 5.22, compared to 3.88 for the others. To me, the most surprising part of this is that it so thoroughly beat relative frequency, which had almost no correlation with perceived accuracy of the sentences. It seems that as long as you can remember the last time a person showered, it doesn’t matter as far as the present simple is concerned that most people do it far more.
It is worth pointing out that the three survey items whose means fall below 4 on the 7-point Likert scale had longer periods since the last occurrence of the action they describe. Bob hadn’t been to the Grand Canyon in 4 years, Cromagnastan hadn’t invaded a country since 1910, and Lauren hadn’t had a seizure since she was a teenager (5 years ago). Only Penny and her poor pets placed high on the accuracy of the present simple despite having no recent occurrences. I would attribute this to markedness, but as we have seen, markedness was a poor predictor of the perceived accuracy of the present simple overall. I am left with no answers on this particular point, except that answers for this item were correlated at 0.54 with answers on the Fluffy the cat question. With Penny as the exception, the surest sign of the acceptability of the present simple to describe a behavior is that it occurred at least once in the past year.
The stereotypically true scenarios/sentences were only the ones about the park and the toothbrushing. I am open to the idea that I should have made more “José eats Mexican food”-style scenarios. Perhaps someone else will study the correlation between present simple acceptability and conformity to stereotype in more detail. I know the greatest book ever by Daniel Kahneman does support the idea that people find propositions that conform to stereotype easier to believe. For example, if I had a scenario like this “José was born in Mexico. In 2011 and 2016, he performed Mariachi music in concert”, a sentence like “José performs Mariachi” would be judged more accurate than if I had given the same situation starting with “Joe was born in Colorado.”
The scenarios that included variation with similar actions, in my case the Grand Canyon, pets, seizures, showers, raccoons, and robberies, may also deserve more detailed study. Some of mine are probably open to debate – I myself am not sure a period of good health constitutes “variation” in a pattern of seizures in the same sense that taking a vacation in Las Vegas is “variation” in a pattern of taking vacations at the Grand Canyon. In looking further into this question, other considerations might be whether one or many behaviors interpose the one in question, what dimensions they are similar in (purpose, length, intent, etc.), and the ratio in frequencies between the behavior in question and the interposing behaviors. As “variation” as I defined it was a somewhat strong indicator of which sentences were deemed accurate, it is probably worth a second look, much like “objective frequency” and “stereotyping”.
As I found when I surveyed people on other grammar points, people tended to be quite confident in their judgments. Out of 410 total answers (41 respondents x 10 questions), 44% were either 1 or 7, and 32% of all answers were 7. The qualitative section also rang with bullish confidence in participants’ judgments. Here are a few answers for the section asking participants to explain their own criteria.
I mainly considered the degree to which the people described were acting of their own volition, and the frequency with which the behavior occurred in relation to my preconceived notions of how frequently the behavior should occur.
Whether it seems like a consistently occurring event or something that is always true. Whether it can be seen as an atomised/whole action. In some cases it was more what was missing to make the sentence make sense that made me disagree e.g. “if it said ‘occasionally’ I’d have no doubt, but without it there’s ambiguity”. Fun exercise.
Expectations as to how often things generally occur (showering, invading, for instance), intervals between recurring actions, expectation that actions will recur in the future, regularity of intervals [Note: even this person’s answers didn’t seem to place high value on relative frequency]
Passiveness/activeness of the person involved
I looked for evidence of a habit that is likely to continue. Jenny has a plan to shower even though it is not frequent. The feral cat has not changed his residence and we know that robberies are likely to continue.
Disagree: Cromagnastan has no plan to invade and what they have done is infrequent without signs of a habit. The cat-killer is similar to Cromagnastan–not enough evidence that it will continue. Bob only went to the Grand Canyon twice, so that is not habitual. Same with Lauren’s seizures.
Overall frequency of events, time between occurrences,
In reading the various scenarios presented in this survey, I considered such criteria as the frequency of the action in question and whether it still occurs regularly in the present – I would thus have amended some of the given present simple sentences (with which we were asked to agree or disagree on a cline) with qualifying adverbs of frequency in some cases, or in other cases, I would have changed the present simple to the present perfect. For me, in general, the use of the present simple in a single decontextualized sentence (though context is given in detail in this survey prior to each sentence) tends to imply that the action of the present simple verb takes place always or regularly. I also feel that ‘context’ and ‘co-text’ are ‘everything’ in communication and language. A very interesting survey. I look forward to reading the results in due course.
With people, time is relatively limited, so I’d say most of these questions are binary. With the country invasions, it becomes a challenge as regimes change. It is safe to say that Country X is different now than it was 100+ years ago, so unless speaking historically, it may not be fair/productive to consider them invaders. Though it may technically be said that they invade, it should in fairness be qualified.
Last, of probable relevance, survey participants described themselves almost unanimously as English teaching professionals and native speakers of English. This says quite a bit about the demographic that constitutes my Twitter followers. Only two respondents identified themselves as native speakers of another language (Dutch and Spanish), and one of these identified himself/herself as an English teacher as well. This may be relevant because English teachers, more than most people, are familiar with the usual terms with which the present simple is commonly introduced, and you can see this in the way that they describe their judgment criteria – they use terms like “habitual” and “regularly occurring” perhaps more than a random selection of Americans would. It is possible that this biased them toward considering these factors more important in making their judgments.
I’m encouraged that so many people considered this an interesting enough topic to take the time to click through my questions. I believe language teachers in general can’t afford to take grammar textbooks concepts of “simplicity” at face value – as we all know for modals, countability, and of course articles. The present simple as I understand it seems to appeal to some intuitive process by which humans see the world – a pattern of agent and action occurring with enough frequency/characteristicness/consistency to warrant a kind of permanent labelling – but people clearly disagree on what contexts justify its use, for reasons that be very individual (how often should one visit a great-uncle?). Furthermore, English learners, even if they have the same kind of “labelling” module that fluent English speakers seem to have somewhere in their brains, have no reason to think that pairing a subject with a verb in its base form or with an added “s” would be the way to express it (or for that matter, why English speakers consider a habit all of whose constituent actions occured in the past the “present”). Of course, many of us prefer not to use grammar textbooks or to organize our syllabi based on grammar points at all, and in those case we don’t need to worry that one or another grammar point is being introduced too early. We might still want to keep in mind that the process by which information about something or someone can get vacuumed up and compressed into a very short present simple sentence is somewhat opaque even to trained linguists (as many of my surveyees are), and there is no reason to expect the present simple to be applied naturally sooner than many grammar points that are traditionally considered more complex.
Thanks for reading. Here is a dog watching some ducks.