Sunday, June 21, 2015

New website

Visit my new website:  www.jeffreypunske.com

Saturday, July 19, 2014

"Word Crimes"


I haven’t posted in a long while and I really shouldn’t be posting now. Things are a bit busy and chaotic as I prepare for my move to Carbondale. My house currently looks like a box factory threw up on a tornado site. I still have various phone calls to make. I have classes to plan and prepare and a half-million research projects on the back-burner right now. In short, I should be working on just about anything else. So why am I writing?

Weird Al.

I’m sure you’ve seen it by this point, that new Weird Al video where he lists the assorted assaults on the English language that he finds so abhorrent. And if, by some miracle, you have not: https://www.youtube.com/watch?v=8Gv0H-vPoDc

Many other linguists have posted responses to this and I don’t want to repeat their discussion too much. I do highly recommend Lauren Squires discussion and pedgogical ideas: http://languagelog.ldc.upenn.edu/nll/?p=13521

Instead, I want to use this as an opportunity to reflect a bit on my time at Kutztown as I prepare to leave and along the way discuss why I find this song so dangerous and damaging.

At Kutztown, I had two main jobs: I taught introduction to linguistics primarily to students preparing to be high school English teachers and I taught freshman writing (six whole sections!). These duties really highlighted why it is so critical that linguists aggressively respond to the Weird Al’s of the world.

Linguists often talk about the differences between descriptive and prescriptive rules-- I’m not going to go into the whole distinction here, if you are unfamiliar, there are plenty of resources out there. To me, this distinction had always been largely academic. I’ve always been a confident writer (whether or not I should be is a totally separate issue). So I was personally never really troubled by many of the prescriptive rules. I may have also been lucky to have reasonable English teachers who understood that a particular peeve is not a hard and fast rule. In my experience, many professional linguists have similar backgrounds. It does take a certain confidence to 1) tell your family you’ve decided to be a linguist 2) enter a PhD program 3) finish said program. So, by and large, linguists view these rules with a bit of joyful, comic disdain. However, we often fail to recognize the damage that these rules can cause.

Kutztown helped put that into perspective for me.

There is a lot to be said about the preparation students get for college level writing in K-12. (The rise of testing couple with its content-free 5 paragraph essay has done major damage to a generation of students.) But, I got to see first hand the real damage that these cause. They convince students that 1) don’t understand the language well enough to write it effectively 2) that there is something deficient about their intelligence which prevents them from grasping the logic behind whatever rule.

With virtually no exceptions, the real issue is that the student’s native understanding of the language is more complicated and complete than whatever dumb prescriptive rule the instructor happens to be pressing. Language is systematic, prescriptivism is not.

I’m not going to look at the particular peeves that Weird Al hit on, but at a broader spectrum of “grammar rules”.

“Rules” such as these are never rules.

I want to start with English “they”. We typically describe English “they” as the third person plural pronoun. As such, sentences like the following are typically treated as errors:

Someone left their bag in the lobby, they may claim it at the lost and found.

The argument is that “they” in the above sentence is corefering with someone which is singular, so there is a mismatch between number that is unacceptable.

This, however, is based on a faulty assumption of what English “they” actually means. English “they” is probably more accurately described as a default pronoun-- we use it when our other pronouns don’t quite fit. That happens to be in the plural, it also happens to be in singular contexts where the gender is unknown or obscured.

What’s more, this use is not novel. Such use can be found throughout the English canon (Chaucer, Shakespeare, King James Bible, etc.). Style guides in the early 1900’s describe singular use of “they” as “old-fashioned” and suggest using “he” in its place. The singular use likely dates back all the way to English borrowing “they” from Old Norse.

However, most pedants view this use of “they” as a new desecration to the language brought on by lazy, inconsiderate youths who won’t get off their lawn-- and it is treated as an such.

Even instructors legitimately trying to help, end up framing the correction in a way that suggests that writer is deficient in some way. It is not unreasonable for a teacher to say that, in a particular style or genre of writing, singular “they” is not acceptable. But note how that ban is framed; it acknowledges that the ban is stylistic, not grammatical.

Telling student that their use of “they” is “incorrect” because of “agreement” is much more problematic. It is saying “you are too stupid to understand numbers” or “you have learned your language wrong”. Obviously, neither is true. The student fully grasps all of the uses of “they” that English provides-- the instructor is the one pushing deficient knowledge.

It is also critical to recall that this knowledge is subconscious. The instructor knows the true meaning of “they”, they (see what I did there) just don’t know that they know it.

More to the point, throwing half-cocked, poorly framed rules as students leads to greater confusion and critical damages a student’s confidence with writing.

Students must learn to diagnose problems and correct them on their own. Throwing unnecessary terminology and a stew of partially understood grammatical concepts at them, helps no one. The passive is maligned and misunderstood by students and writing instructors alike. See: http://chronicle.com/article/50-Years-of-Stupid-Grammar/25497
We’ve really got two major problems with the prescriptive ban on the passive. First, too many people (Strunk and White) cannot properly identify the passive. How can students be expected to know what to avoid when the style guide identifies a cacophony of grammatical constructions as the passive?

Last term, I had a student ask me to help him with a note he got from another professor. The student had written the following sentence: 

His arm was draped over the side of the couch.

The professor had circled the sentence and wrote with the note “passive”. This sentence is ambiguous. There is a passive reading and a stative reading. To get the passive reading we need to imagine a scenario wherein some force has taken his arm from a non-draped position to a draped position. This is a possible reading—but probably not a very likely reading. But critically it requires an event to take place. It was pretty clear from context that the sentence was intended as a stative. The student was describing an image from a film. Everything in the relevant paragraph was a description of the scene. There were no movements. No events.

In other words, that sentence ain’t a passive! And it is completely clear from context. And we can shift the context to clarify the ambiguity.

Consider:

Scenario A:

Watson: It was a suicide.

Sherlock: This was no suicide, Watson. His arm was draped over the couch to make it appear as one.

Scenario B:

Lestrade: The scene was quite messy. We found the body on the couch. His arm was draped
over the side
.

----

Similar misunderstandings abound about the passive. Strunk and White batted only 1 for 4 on their identification! That’s 25%. That’s F-level work right there...

And second, the passive totally has legitimate uses!

We can use the passive when the subject can be generally understood by our encyclopedic knowledge of the verb and we want to focus on the object:

The police arrested John yesterday. (Meh... the police are the types of people that do arresting.)

John was arrested yesterday. (Dude, we are probably telling a story about John anyway)

We can use the passive when the subject is unknown and introducing a generic subject would
cause us to focus on the wrong element. Imagine you are reading a crime novel. Which opening
sentence would draw you in?

John was murdered yesterday.

Someone murdered John yesterday.

If you answered the second one, well, I don’t know. I don’t believe you. At all.

We can also use the passive when the subject is basically irrelevant (because it is understood by context). We see this a lot in descriptions of experiments or other similar contexts.

Heat was applied evenly to the glaze for 30 minutes.

The experimenters applied heat evenly to the glaze for 30 minutes.

There is just a small sample of instances where the passive might be preferred over the active.


Let me also be clear, I sometimes correct the passive in student work. Young writers tend to overuse the passive. It is perfectly acceptable to say, “hey, a passive isn’t appropriate here”. And this draws me to my larger point. Good writing takes a lot of hard work and a lot of practice.

A few simple magic grammar rules can’t suddenly transform a novice writer into a brilliant one. Rather, it takes years of blood, sweat, and tears from both the pupil and the instructors. To suggest otherwise is offensive.

Further, to reflexively deny a writer access to one of the tools in the writer’s toolbox is not sage advice.

Moving even further away from Weird Al, I want to turn to the recent cottage industry of machine-driven writing coaches (which are made all the more scary by the fact that testing agencies use machines as part of their “evaluation” of student essays). Here I examine the “Writer’s Diet Test”. I’m not going to provide a link, because, you know they’re evil.

Basically, you copy some text in and it gives you a rating of “how fat” your text is and also provides some suggestions.

And let’s be honest, we are probably not that far from the day when some testing agency integrates something like this into its evaluation process. So, what’s wrong with it? First, as your disclaimer notes, judging the overall quality of writing is not something that can be done by a machine. I certainly have to say that I agree. Let’s examine some of the overall results (based on a random 1000 words) of your test.


Hamlet’s “To be or not to be” soliloquy: fit and trim (how gracious of you)


MLK’s “I Have a Dream” Speech: needs toning


The Gettysburg Address: flabby


Churchill’s “Blood, Sweat, and Tears” Speech: flabby


The first 1000 words of this blog: flabby

Awesome. I’ll take any comparison to Abraham Lincoln I can get (except maybe in the looks department).

I have to say, it is one of the dumbest parsers I’ve seen in a long time. From what I can gather, the test is supposed to highlight every example of an adjective or adverb and examples of nominalizations and some other junk. To test it out, I input the following sentence:

The gilded lily is found throughout the impoverished nation.


Now, I know that the test is not calibrated to handle examples of less than 100 words. So, I really don’t care about the overall score. I’m interested in how the test treats the various words here. Also, I’m not saying that this sentence is an example of great writing.

The results are not pretty:

The test highlighted lily as a problematic adjective or adverb. This is, of course, patently absurd. Lily is a noun. Why is the parser confused? Apparently, it is coded to treat any word that ends in –ly as an adverb. Seriously? They didn’t even bother to give it a lexicon of basic words that might trip it up?


There’s a similar problem with nation. Nation is highlighted as a problematic nominalization. Obviously, because the word is derived from our verb na to... oh I don’t know, you make something up….

Neither gilded nor impoverished are identified as adjectives—which they clearly are.Identifying these forms correctly would not be a particularly difficult computational problem.

So what does all of this have to do with Weird Al?

Weird Al is part of a long lineage of folks who think that their particular peeve and/or fundamental misunderstanding about how the language works out to be codified and used to evaluate the intelligence and ability of others. It is wrong and it is damaging. And it is always based on an incomplete analysis of the language.

There is a lot more than could/should be said about all of this. There are important interactions of privilege and class that prescriptivism is built on, but I need to get back to packing.

Feel free to comment below.

Edit: A clarification to the texts used with the writer's test.

Wednesday, October 9, 2013

Been too long

As you may have noticed there's been a long lag in posts here.  Things started to get busy and it just fell out of my routine.  I'll try to pick things up again.

A lot has changed since my last post.  I, sadly, departed the University of Oklahoma and am now at Kutztown University in Pennsylvania.  I'm definitely keeping myself pretty busy out here.

Some exciting developments: The October 2013 issue of Lingua contains a paper of mine.  Scott Jackson and I have a paper coming out in the next issue of Linguistic Analysis. I'll be headed up to NELS as an attendee, giving a talk at Wheaton college, and then presenting at the LSA in January. I'll also try to be more active here.

Wednesday, April 3, 2013

A quote too good not to share

The last part of instructions for an assignment (from a friend):

This is the assignment. There's a lot of room to walk around in it. If you think you might be lost, walk back to me and check in. The path will still be there.

Tuesday, April 2, 2013

Independent undergraduate research

I posted my reaction to Lasnik's article on teaching syntax a few days ago.  This is another teaching related post, but on a slightly different topic.

Linguistics at OU is both massive and tiny.  We are the second largest program in our department in terms of undergraduate majors, but we are the smallest in terms of faculty.  In the fall, it was just Marcia and I (Dylan was on sabbatical).  Now, it is just Dylan and I (with Marcia on sabbatical).  This means that the students get a LOT of contact with us.  I have several students who have taken 4 courses with me this year.  And a majority of my students have taken at least 3 with me. Of course, that means that they get to know each other really well too. That means that we get to know each other pretty well, and we get pretty comfortable with each other.  I think that's great.  It gives the classes a lot of energy.

Our small number also means that students with interest in topics in linguistics not explicitly covered in our courses, only have a handful of folks to turn to.  I have an open door policy and, apparently, a pretty interesting bookshelf.  So students are coming by all the time to chat about something they learned in psychology/Chinese/French/Cherokee/German/whatever or ask to borrow a book (typically, "just for fun").  After encouraging them to find at least *a* hobby outside of linguistics, I try to point them to ways to apply whatever they learned to some course concept or linguistics in general.  I always encourage them to start to pursue the topic independently.

And that's what I'd like to focus on today, building independent research with undergraduates.

I get questions fairly frequently on a broad range of linguistics topics. Sometimes, it is on something I know a lot about. Other times, I don't know much at all. I know I frustrate some students because I don't know EVERYTHING about linguistics. I always try to give some answer. At the very least, I always know enough to know where to look. Recently, this method has evolved to me giving the student an accessible paper that outlines the basics of the topic and scheduling a meeting for them in a week or so.

At this meeting, I try to help them interpret the paper, if they need it.  But I emphasize that they are now the expert on campus on this topic.  We cover a little bit on how to find more papers and how to use the papers we have to follow up on topics.  But I really want them to walk away with the idea that they don't need me to pursue their interests -- I'm there for them for sure -- but, they've got the skills to tackle a lot of these questions themselves.

I'm lucky to get to work with such an engaged and thoughtful group of students. Is there a better feeling than when your students get to teach you?


Sunday, March 31, 2013

Teaching Syntax/Teaching Linguistics

This post is a reaction/response to Lasnik's new article "Teaching introductory graduate syntax".  Mostly, it is a full-throated endorsement of everything Lasnik has to say with some discussion about how the basic ideas Lasnik presents can and should be extended to teaching linguistics more broadly and some of reflections on my own.  I highly recommend Lasnik's article.  It is a very thoughtful piece and I'm sure that there are things to be drawn from it regardless of your particular area. I know that reading this article will improve my teaching.

Lasnik presents emphasizes three major points:  1) Tracing the development and history of the framework-- not just teaching the latest architecture; 2) timely and thoughtful responses to written work, coupled with the opportunity for students to redo assignments; 3) the importance of an interactive classroom.

I will discuss each of these in some detail.  I will reflect back on my own graduate education, how I feel that  these insights apply more broadly, and how these points relate to my own experience teaching an undergraduate Syntax II class at OU.

1)  Tracing the development and history of the framework

To me, this is one of the most essential points that Lasnik makes in this article.  In many ways, the latest innovations in syntactic theory only make sense if you are grounded in what preceded them. More fundamentally, there are a number of early generative and pre-generative discussions and analyses that have direct application to the work being done today.  It simply makes no sense to ignore the work of our intellectual ancestors. I have spoken to advanced graduate students who have been completely unaware of major historical works that bear directly on their analyses-- that shouldn't be.  Training students in the development of the field is not only essential for their understanding of the modern frameworks, but also for allowing them to understand and discuss the particulars of previous analyses in slightly different frameworks.

Thinking back on my graduate education, I was quite blessed to work with Andrew Carnie* for the bulk of my early syntax education.  Andrew believes not only in teaching the history of the field but looking at other related frameworks.  I truly believe this made me a better researcher and scholar.  In my own research, I seek out the relevant pre-generative and early generative work.  In every case, this has improved the scholarship.  While I certainly have my own particular failings, Andrew Carnie's early guidance certainly made me a better scholar.

The course I am presently teaching, Syntax II, is an interesting course for a number of reasons.  It is not a typical offering at OU.  This may well be the only time it is offered for a decade.  It is an undergraduate course, which changes some of the emphasis and dynamics of the course.  Graduate courses are designed to train students to be professionals in the field.  Undergraduate courses are designed to train students in ways of thinking about and analyzing problems from a linguistically informed perspective-- but with a more general goal in mind.  Most undergraduate linguistics majors will not go on to pursue a career in linguistics (given the job market, that is a very good thing). So, the goals are different.

Students enrolled in my Syntax II course, are required to have completed Syntax.  Unlike Lasnik, I have a very good sense of the background that my students came in with.  All of them learned syntax out of Andrew Carnie's textbook-- a book that I essentially have memorized.  However, there still are notable difference in students' retention and understanding of the previous material. Many students completed the course in the Fall; for others, it has been a couple of years. At the end of the day, I was faced with a similar problem.

In the Fall term, while planning this course I asked several students who were interested in the course what particular topics they would be most interested in covering.  There were a number of different approaches I was considering for the class.  We could treat the class as a simple continuation of Syntax, complete the sections of Carnie that were not discussed in the previous versions of the course, and add in more topics.  Or, we could introduce the Minimalist Program.  Or, a mix of the two.  Overwhelmingly, the students wanted to learn MP.  But there was a vocal minority that preferred the advanced topics.

After a lot of consideration, I decided to approach the class in a unique way.  Since I knew the basic background all the students would come in with, I wanted to play off of that a bit.  Rather than just teaching pure MP, I decided to let them make the decision.  The class has become a comparison of Carnie-style X'-based, P&P syntax and Adger-style BPS MP syntax.  We cover advanced topics and we cover MP.  However, the particular analysis adopted is up to the students. At times, the class has struggled with the idea that I am not teaching them the "right" answer-- but, overall they have embraced it.  Most assignments involve problems that must be done in one approach or the other and problems where the students must choose an approach and explain why.  At this point I have a 1/3 of the class that has a clear preference for MP-style analyses, 1/3 that prefer X'-style analyses, and 1/3 that is still trying to decide.  That is the best outcome I could hope for.

This is a very long-winded way of saying that I opened the class with a two-week discussion of the historical development of syntax.  The only way to understand why the GB to MP shift happened is to understand what came before it.  This foundation has greatly helped the students evaluate the claims we discuss in the class.

2) Timely and thoughtful responses to written work, coupled with the opportunity for students to redo assignments

I think the first point is obvious, and I would encourage you to read what Lasnik says about it.  I will not rehash his points exactly.  I have a strict 7-day return policy in all of my courses.  Preferably, I will get work back to my students by the next course period; but with 3 courses that is not always possible.  But all written work is returned within a week.

I really liked Lasnik's discussion of redoes. This is something that I will be adopting in all of my future advanced-level courses.  I've had a forced redo policy for some time.  If an assignment shows a very poor understanding, I will have the student redo it after discussion and grade the final version.  Students are typically capped at a B under this policy.  However, I much prefer Lasnik's system.

Lasnik allows all students to redo their assignments.  The redone assignment's grade replaces the original.  The critical thing that Lasnik does is that he requires students to write-up a discussion of the changes that they made.  They must explain why they made the changes and what they know or understand now that they didn't when they first completed the assignment.  This is brilliant.

The most important aspect of any class is that students understand the material.  Whether or not they get it on the very first pass, shouldn't be relevant.  It also prepares them (in a subtle way) for dealing with comments and revisions which is something that drives an academic career.  Yes, it produces more work for the instructor, but the results would definitely be worth it.

I would also hope that it would allow students to take more risks in suggesting new ideas and to reflect and revisit problems-- both extremely positive things.

3) The importance of an interactive classroom

Much like the homework discussion, I don't have much to add other than Lasnik got it exactly right.  The classroom needs to be place where students feel comfortable asking for clarification, presenting troubling data and challenging the analyses being presented.  Many of my best classes (across multiple subjects) have been the ones that start with an early question from a student and build from there.  Every question need not be answered on the spot, but it is critical that students feel involved in their own learning.

Embrace the unexpected.  Pursue whatever analysis your students come up with.  It makes the class more fun, for you and for the students, and it really helps sharpen the tools of analysis and hypothesis building.   It can lead to really great discussions-- and honestly, help develop your own research.

I really love building our discussions collaboratively in class. It does make classes harder to prep.  But so much more fun in real life.  Sometimes, the students will astound you.

An anecdote 

A few weeks ago, I presented my students with some very tricky verb particle data.  I expected to spend most of the hour pursuing dead-ends.  After discussing the basics of the problem, I asked the students if they had any initial hypotheses about how to deal with the data.

A student raised her hand.  I had already preplanned the 5 or so potential answers I was expecting to get and some basic talking points to build from each.  I call on her.

Student: "Maybe... {the solution I as hoping to get us to today}"

Once I got done picking my jaw up off the floor, I said, "Yes.  Exactly. That's great.  Now, how did we get there?"

And the class then proceeded more or less as planned.  We still discussed all the analyses that wouldn't work.  We went through all the arguments and built the analysis as planned.  But, I also got to praise a student for some truly outstanding insights.
----

Lasnik's article has a lot more discussion of all of these points.  I highly recommend it to everyone.  It really isn't about teaching syntax-- or even teaching linguistics-- it is about teaching.

I'm really looking forward to the rest of this e-series from Language. I'd also love to hear your thoughts in the comments.

References:

Lasnik, Howard. 2013. Teaching introductory graduate syntax. Language 89.1: e11-e17 (online)

Notes:
* I do not mean to suggest that the other Arizona faculty do not take such an approach.  I just happened to take my foundational courses from Andrew.

Thursday, March 28, 2013

busy time of year

I've been much too quiet here lately.  I have a couple of new posts that are in the works, but I'm not satisfied enough with them for public consumption yet.  One in discussion of Lasnik's new article about teaching syntax and the other will have to be a mystery for now.

Though, I haven't been active here, I've had a very productive March.  Scott Jackson and I completed and submitted a paper on compounding and I completed are resubmitted revisions on a paper on the syntactic derivation of morphological regularity.  I'll have more on those as well as they weave their way through peer-review.

Hopefully, I can snag some time and get those backlogged posts up soon.