One of the lovely parts of a PhD defence in the Netherlands is that it’s often accompanied by a small symposium. The idea behind it is that external examiners don’t travel a long distance just for the defence, which is over in an hour, and diner afterwards. In my case the topic was action formation, obviously, but it consisted of varying perspectives that showed the breadth of views that CA has to contribute.

As with any defence symposium, I had the honour of giving the first talk. But before I could start it was clear that I had made a slight error: I had seriously underestimated the number of people who would attend. There were well over thirty people and of course I only had twenty handouts. It’s a problem well known to many researchers in CA, and it happens all the time, but not a great way to start the day you have to defend your dissertation.

Back to the Future
I took the opportunity to present some new work. Typically a graduating PhD will talk about their dissertation. I remember Trevor Benjamin giving a very interesting talk about how little we know about repair, which is all the more surprising considering repair is such a fundamental part of social interaction. But I went a different path, and picked up a small topic I had worked on as a visiting graduate student at UCLA: past tense.

From Functional and Cognitive Linguistics we know that tense is often not temporal: for example, we use the present tense to talk about the past. My aim in this talk was to present some initial findings about what participants do when they use the past tense to talk about future events—the talk was entitled Back to the Future for obvious reasons. While this use of the past tense isn’t very frequent, I had collected thirteen cases and I argued that there was a systematicity in how they were used: speakers would use it when asking questions to which they should know the answer, because they had been told by their interlocutor.

These cases showed in a lovely way that knowledge is first and foremost social: when you violate your epistemic responsibilities, when you should know but don’t, you have to in some way show that you’re aware of this violation. In other words, you have to show that you know that you should know the answer. And indeed, I demonstrated that when people ask questions to which they should know the answer without a practice that shows this awareness, they are sanctioned for it.

Giving advice in interaction
Up next was Betty Couper-Kuhlen who presented collaborative work with Sandy Thompson on suggestions. While they have been studied a lot in institutional settings, such as psychotherapy where the goal is to offer advice, suggestions have been somewhat ignored in casual conversation. She neatly demonstrated that there are various ways in which people can do suggestions, and that these suggestions can be elicited or volunteered.

The form of these suggestions provide insights into the social relation between the participants. One frequent practice is the why don’t you do X, which Betty argued allows the interlocutor a way out. While it is clearly a conventionalized practice for doing suggestions, and not asking questions, the wh-interrogative can be used by the recipient to provide a reason why indeed he or she won’t follow the suggestion. At the same time it forces the interlocutor to deal with those deontic implications, whereas a practice like You could do X can be more easily ignored. An imperative form, finally, presents the suggestion as an obvious one, which again has very different implications.

Brittish Tutting
After Betty we moved from syntax to phonetics, with Richard Ogden who presented on clicks. As anyone who has ever seen Richard present knows, he can very simply demonstrate the importance of clicks in everyday interaction. They are often not simply affective, where we display our disapproval by providing a few clicks, but have a significant role in turn-taking and turn construction. Clicks can be used to show that you are going to speak or they can be used in the process of a work search. While one might argue that in these cases they arise naturally—you have to open your mouth to produce words—they are understood as meaningful.

The weirdest cases are of course where multiple clicks are produced in close succession. These can be the typical dental clicks, but also lateral clicks, you know, the ones we use for horses. While they lack a clear semantic meaning, they can be used by participants in a way that shows a clear intersubjective understanding of their function. Clicks present a world of meaning, and one that Richard keeps revealing bit by bit.

Action formation in surgery
In the final talk we moved away from conversation almost entirely, as my daily advisor, Mike Huiskes, presented on interaction during orthopaedic surgery. This was collaborative work with Lotte van Burgsteden, currently a PhD student at the University of Wageningen. Mike showed that while there is very little talk during surgery, the physicians are continuously monitoring each other for interpretable behaviour. A surgeon who can be seen to be doing looking, will be understood to be doing looking for a reason, which can subsequently be interpreted by others as for example an assessment. As Goffman argued that you are always seen to be taking a line through talk, so Mike showed that even if you’re not talking, you can be seen to be taking a line.

Through his talk he raised the question, inspired by recent work by Nick Enfield and particularly Jack Sidnell, about what action is? Is social interaction about action or is it primarily about behaviour? The distinction between the two is about whether or not specific intentions are explicitly ascribed, and as Mike likes to repeat: action is conduct under a description. A view like this clearly clashed with long-held views in CA where Sacks argued that it should be a systematic science of social action. But it’s a very interesting debate that certainly will lead to inspiring new ideas about social interaction.

These four talks together provide a wide view of CA and Interactional Linguistics in particular. They demonstrate that much of what linguists have long assumed about language and particularly language use is not as straightforward as might seem. Linguistic structures are produced and understood in interaction, and they therefore cannot be exhaustively studied isolated from their interactional environment. And indeed, as was recently also argued in a special issue of Research on Language and Social Interaction, we need to study embodied interaction to understand grammar. It was all in all an inspiring and educational start to a day that of course, in the end, was just about me getting an oversized degree.

The Student Union at the University of Groningen recently published the results of a survey on stress among students. The results: two thirds of the students in Groningen experience stress regularly or even all the time. That is a lot, and the consequences could be serious. I remember having sleep issues during my undergraduate studies as a result of stress, which also affected my performance in class. If two thirds of students experience similar symptoms, we clearly have a problem that needs to be addressed.

The study is a one-shot survey, so it is unclear whether the current student population is more or less stressed than previous generations. But there are reasons to assume that things have gotten worse. Since the recession in 2008 the Dutch government has been trying hard to balance the budget, and it has similarly forced semi-public institutions like universities to get their finances in order. One of the major causes of budget deficiencies are students who take too long to graduate; they should finish their BA in three years and their MA in one or two, depending on the type of degree.

But few students actually manage this. It took me a six! years to get an undergraduate degree. The university is paid for only three of those years. It works as follows. Tuitions in the Netherlands are low, because the government uses taxes to cover most of the actual costs of a university program. But universities are not paid for the number of students they have, they are paid for the number of students that graduate. So the university will make the same amount of money on a student's education, independent of whether a students takes three or six years. But if a student needs six, the university does have to provide facilities for that student such as classes and library access for six years. So students cost money if they don't graduate on time.

In order to fix this, universities have implemented a number of measures that as a side-effect increase stress. I want to deal with one in particular. Many classes I TA have forced student participation. When I started my undergraduate degree, virtually all classes and all homework was optional. As long as you took the mid-terms you could take the exam. Nowadays, students have to be in class, have to participate, and have to hand in their homework. To be clear, I don't grade the homework; it is just a means of forcing participation. In this way the university can make sure that students stay on track and know the material when the exam comes around.

Forced participation is not really a good solution though. It is easy for the professors and TAs; we know students do the homework and engage with the course material. But students do not learn much from it beyond those materials. We are basically teaching to the test. Getting a higher education is not just about learning more complicated stuff, it's about personal growth as well. Students need to be prepared not for a specific career, but for a broad future and part of that is engaged citizenry: they need to become responsible adults. Mandatory homework just for the sake of having homework does not accomplish that.

In fact, if anything it has the opposite effect. Studies have often shown that forced learning is not as effective as voluntary learning. It can make the material less fun or even annoying, and that will mean that students learn less. Even students who could have been enthousiastic are just going through the motions. Learning is supposed to be fun, but most of us feel that studying is not. And of course, since students do the work because they have to, not because they see the benefit, they do not get the opportunity to learn such skills as responsibility or self-discipline.

The negative effects of mandatory participations are one of the main reasons the political organization for young conservatives (JOVD) wants to get rid of it. Students should be treated as adults. While there will be plenty who have a hard time making the switch from high school and living at home to college and living on one's own, we should provide support for those students so they can grow instead of keeping them small.

But the JOVD offers an additional point of critique: students are clients of the university and so the university should accomodate them. And this seems a weird position to me. Sure, students—or their parents—pay tuition to attend college. But through taxes society still pays well over 80% of the students' fees. So by that logic colleges should cater to society first and to students second.

But even if we were to take the perspective of the JOVD—all over the world plenty of people pay all their tuition fees themselves—that does not mean universities should just do what students want. Students pay for an education, they want to learn. They are not buying a degree. This means that they should expect universities to do their best to educate them, and that requires a lot of student participation. I can only adequately teach if students come to class willing to learn. If they sit back and just expect that we somehow magically transfer our knowledge and skills to them, they will never get anywhere.

You can treat higher education according to free market principles. But every market is different. Education is a service, not a product, and you cannot treat is as the latter. So yes, let us by all means get rid of forced learning, but when we reorganize the university, let's do so in a way that makes sense. Student-centered learning means doing our best, together with the student, to provide them with a decent education. It is not and cannot possibly be a one-way street. 

I was recently on the train, coming back to Groningen from giving a guest lecture at Radboud University in Nijmegen, when I got a phone call from my credit card company; whether I was interested in accident insurance. It was a Friday, I was tired, and the first month would be a trial, so I said fine. If nothing else it would give me the chance to look over the policy and see what it would actually get me. Turns out it was a good thing I did, because reading it gave me a good laugh. In the clause dealing with when my next of kin would get money in case I died, it said the following:

Did you, or another insured party, get into an accident and die as a result of it? 
('Heeft u, of een andere verzekerde, een ongeval gehad waaraan u bent overleden?')

That's right, the policy asks the reader whether s/he is dead. I clearly cannot answer yes, as it is particularly hard to read the policy from beyond the grave. Now maybe the insurance company believes in an afterlife, but let's ignore that option, because even if they did, how am I supposed to tell my family what to do. It's not like they regularly visit a medium.

But note that there is a second fluke in this sentence, and that one is even more funny. The way it is phrased another insured party could have gotten into an accident resulting in my death. Talk about injustice! My hypothetical partner loses control over her vehicle, crashes into a truck, and suddenly I drop dead in front of a group of colleagues during a lecture on another continent. It's just as well that the policy does not cover intent, because otherwise a malicious partner could decide to take my life in some indirect way claim the payment, and not be convicted for either fraude or murder.

The point of the sentence is obvious enough. The insurance company was probably trying to make it easier to read the policy. God knows that most people don't go through the terms and conditions of Apple, Google, Facebook, etc. But they seem to have gone a bit overboard in making an active text. And the lack of care does not bode well for the company itself.

ne of the great things about doing a PhD is that how you spend your time is pretty much up to you, at least, if you don’t have to run specific sets of experiments or observational studies of course. But since my data had already been recorded by the time I started, everything was up to me. That’s great, because it means you can easily spend some time at different research institutes, working with and learning from other scholars. And lucky me, I got to spend two terms at the Mecca of Conversation Analysis: the University of California Los Angeles.

As much as I’ve come to appreciate the research environment in Groningen over the last couple of years, it cannot compare to UCLA—and probably other great research institutes. In Groningen there are only a few PhD candidates in our research group, and most of them don’t work on their PhD fulltime. They have actual jobs, and write a dissertation on the side; the faculty in the meantime has an enormous teaching load, and so has little time for research; and we lack the funding to regularly invite famous scholars for colloquia or workshops. People also generally don’t really look beyond their own niche; not only do linguists not mingle with for example sociologists, but excluding one short monthly lunch, the different fields of linguistics don’t even meet.

UCLA was research heaven in that sense. Not only did I get the opportunity to take classes from some of the most influential scholars in the field—Tanya Stivers, John Heritage, Steven Clayman, Charles Goodwin, and so forth—but it was far easier to go to lectures on topics I would never have seen in Groningen. It’s great to be able to expand one’s horizons with talks in anthropology, nutrition, or geography. I wouldn’t say I’ve become a Huomo Universalis, but I have definitely broadened my mind.

It’s also weird how productive you become in such an environment. I’m a pretty quick writer, partly because I can get an analysis down quickly, but also because I have a decent work discipline. And still the amount of research and writing I got done in just a few months is so much more than what I would normally have done in Groningen, while still having time to explore everything the west coast of the USA has to offer. Just in the fall I managed to write a complete paper, from idea to analysis, in a matter of months, and having finished that I resubmitted a second paper, and wrote a third, this time even collecting and transcribing my own data.

One reason is of course that you want to impress the researchers who read your work. It’s a great opportunity to show to top scholars what you’re capable of. But it’s more than that. The frequent conversations, discussions, and feedback sessions provide you with so many ideas, that you have enough energy to just keep going. I still have plenty of ideas for projects I want to work on: For example, how do we use the past tense in conversation? It’s definitely not only for describing past events, but it seems to have a large social and epistemic component as well. Something I probably would not have figured out, had I not written a term paper.

But of course, these trips abroad are about more than just doing research. I’ve always loathed major cities. I never liked Amsterdam or Rotterdam, Paris is massively overrated as far as I’m concerned, and I couldn’t get out of Johannesburg fast enough. The crowds, the filth, the noise: I preferred the quiet life. So moving to Los Angeles seemed crazy: one of the biggest cities in the world, with no decent public transportation, and very little opportunities to walk anywhere. But LA has so much to offer in terms of culture, cuisine, and hiking, that it’s hard to not love the city.

My experience with the US in general was very positive. I got invited to spend New Year’s at a friend’s place in Oregon, and during that break we toured the state a bit; I stayed at a professor’s place for the weekend in Santa Barbara to give a talk and do some whale watching; I crashed at a colleague’s place in Berkeley to visit Oakland and San Francisco: all great experiences and each confirming that Americans are a very hospitable people. My younger brother can also attest to that: when he visited he stayed with one of the guests of the hotel he manages.

It’s always a shame coming home after such a great time, but at least the experience confirms that academia is a great environment to work in. If only The Netherlands had such research driven environments. But I guess you can’t have everything.

The American courts are bound by an enormous set of rules and regulations, typically formulated in a very specific way so as to make it unambiguously clear what lawyers, judges, and witnesses may and may not do. Yet strangely enough, one of the most fundamental rules of courtroom proceedings is not clearly defined. During a direct examination an attorney is not allowed to ask leading questions, except where necessary to develop the witness’ testimony, but what makes a question leading is not really clear. The conventional definition simply states that a leading question suggests the correct answer to the witness or puts up some state of affairs for the witness to merely confirm.

Rules of Evidence
The reasoning behind this rule is pretty straightforward. Only the witness had access to the events that led to the trial, and so only the witness is in a position to provide testimony. If an attorney is allowed to ask leading questions, the witness is simply confirming the testimony offered to him/her. But when does a question suggest an answer to a witness? Some might argue that any yes/no-type question offers a preferred answer, and so attorneys should only be allowed to ask questions that contain words like what, who, or when. But such an approach would make it almost impossible for the attorney to elicit a complete testimony. Moreover, asking a question presupposes that the witness knows the answer, but how would one go about establishing that a witness in fact knows without asking it first?

So it is clear that we need more than word order alone to determine whether a question is leading. We would probably want to consider such aspects as tone of voice, specific phrasing of the question, and of course bodily cues. Combined these factors make a fixed definition impossible, and so it is up to the participants in a trial – the attorneys and the judge – to determine on the fly whether a question is or is not leading.

Non-leading practices
Even lacking a clear set of criteria, attorneys are pretty consistent in how they phrase their questions. While studying the trial State of Florida v. George Zimmerman from 2013 I notices that there are at least two relevant aspects for determining whether a question is leading or not. First, it matters whether the question is preliminary to the witness’ testimony. Preliminary questions deal with such things as the witness’ living situation, and these are frequently leading without being objected to. Second, there is a distinction between questions that address whether the witness has knowledge, ancillary questions, and questions that actually request testimony.

Preliminary questions are not necessary leading, but frequently the witness only has to provide a confirming response. In other words, the attorney puts words into the witness’ mouth. But because the question is preliminary, and typically deals with issues that are not disputed, there is no reason to object to them. To give a few examples: Zimmerman’s attorney asks his character witnesses questions such as the following:

• It’s that friendship that brings you here today. Is that correct?
• I know that you were present during the time that proceeded the event that brings us here today and right after. Correct?
• You acknowledge that you’re a good friend of his. Correct?

With all these kinds of questions the witness only has to confirm a description given by the attorney; they thus clearly are leading questions. Nonetheless the prosecutor does not object, as they are only preliminary questions that establish the relation between the witness and the defendant, a relation that is not disputed.

Knowledge questions
A number of the character witnesses was summoned to testify about the nature of a voice that was hearable in the background of a 911 call. Neighboring residents had called 911 about some screaming they heard, and in the background of on one of the calls there is a voice screaming for help. The defense wanted to establish that this voice belonged to George Zimmerman, a matter that was very much in dispute as the prosecutor had called expert witnesses to testify that it was impossible to determine whose voice it is. Furthermore, it was not necessarily the case that the witnesses called by the defense actually knew whose voice it was, i.e. their knowledge of the matter was not a foregone conclusion. And so instead of simply asking them to confirm that the voice belonged to George Zimmerman, the attorney asked questions like the following:

• Can you identify who’s voice that is?
• Do you have an opinion who’s voice that is?
• Do you know who’s voice that is in the background screaming?

Although the goal of these questions is clear, the attorney is not putting words into the mouth of the witnesses. That is to say, while he obviously wants them to confirm that they recognized the voice, providing a confirming answer would do nothing more than establish that they are able to testify. And so none of these questions is objected to by the prosecutor.

That knowledge, the ability to testify, has to be established can be seen even better in cases where the prosecutor does object. Consider for example the following, abbreviated, dialogue:

Def: You understand now that mister Zimmerman was out of state just shortly after the shooting?
Pro: Objection. That is calling for speculation, unless she knows.
Def: Okay. Do you know where mister Zimmerman was just after the shooting?

The defense attorney is not allowed to ask the witness to confirm some state of affairs she knows about, unless it is first established that she does in fact know.

Consistency
But not all norms are applied as consistently. When dealing with what a witness heard, the attorney is not allowed to ask the question in [1] as it is leading, but the version he subsequently asks in [2] is not problematic:

1. Did they seem to be scream, time, scream, time, scream, time?
2. Did you hear gaps in between the screams?

Both questions suggest that the witness heard multiple screams, and both thus seem equally leading. Nonetheless the prosecutor only objects to the way it is asked in [1]. Yet later on, when the attorney asks whether the witness heard multiple calls for help the prosecutor again objects, and that objection is sustained. It is thus not always clear where we can find the border between leading and non-leading questions.

The issue of leading questions is established very much in the moment. While there are practices that are applied more or less consistently, it is in the end up the attorneys and the judge to determine on the fly which questions are and which questions are not leading. Sometimes it seems that the syntactic design of the question is the problem, but sometimes it is in the choice of words or the non-verbal cues given by an attorney. The participants are thus dealing with a rule that has to be re-established time and again, which makes the trial a bit subjective. But seeing as every case is unique, guidelines are probably the best we have.