Thursday, June 30, 2011
Eating your vegetables: basic definitions and notations
As I am revising the draft of a paper, I am going over a section entitled "Terminology". Like many sections entitled "Basic definitions and notations" that I have written or read in the past, it is boring and seems pointless - until at some point in the paper you suddenly encounter an uncertainty about a precise definition, whose subtler details, heretofore unnoticed, could invalidate correctness, and you suddenly rush back to the definition to check that the details are consistent with your understanding.
The terminology section is like vegetables: a dreary but necessary part of the meal. Is there any way to put a little spice into it, I wonder?
The terminology section is like vegetables: a dreary but necessary part of the meal. Is there any way to put a little spice into it, I wonder?
Visiting Microsoft Research
I am visiting the Theory group at Microsoft Research in Redmond, and this week it has shifted to summer mode: a plethora of visitors at all levels of seniority, talks and research discussions in every corner of our section of the building, and plans for a group weekend hike in the Alpine Lakes Wilderness area. It is a little bit like the French Riviera: quiet in the winter, overflowing with visitors in the summer. The only difference is that the visitors are not here to meet movie celebrities but to work with one another.
The research group is headed by Yuval Peres, formerly of Berkeley and the Hebrew university, and spans a diverse mix of areas: theoretical computer science, theoretical physics, probability and combinatorics, algorithmic game theory. The common theme is what gave the group its name, and collaborations developed there have spurred some of the nicest papers at the intersection of those areas.
The absence of Oded Schramm, virtual Fields medalist (meaning that everyone knew that he would have been awarded a Fields medal if he had not been over 40 years old, and that he would not have been over 40 years old if he had not been Israeli and obligated to spend a few years of his youth doing military service), still leaves a big gap. He died in a mountain accident on Labor day a couple of years ago. Now there is no one to solve people's hard problems with quite the same ease, nor to lead us to ping-pong or fussball after lunch. It's funny how quiet he was in life, yet what a big shadow he leaves.
The research group is headed by Yuval Peres, formerly of Berkeley and the Hebrew university, and spans a diverse mix of areas: theoretical computer science, theoretical physics, probability and combinatorics, algorithmic game theory. The common theme is what gave the group its name, and collaborations developed there have spurred some of the nicest papers at the intersection of those areas.
The absence of Oded Schramm, virtual Fields medalist (meaning that everyone knew that he would have been awarded a Fields medal if he had not been over 40 years old, and that he would not have been over 40 years old if he had not been Israeli and obligated to spend a few years of his youth doing military service), still leaves a big gap. He died in a mountain accident on Labor day a couple of years ago. Now there is no one to solve people's hard problems with quite the same ease, nor to lead us to ping-pong or fussball after lunch. It's funny how quiet he was in life, yet what a big shadow he leaves.
Wednesday, June 29, 2011
Sandpiles
Sandpiles. Just the name suggests recreational science. The science in question is a mix of combinatorics and probability. I believe that the model came from statistical physics, although that does not necessarily mean much in terms of relevance to reality, although I am not sure I know what "reality" means anyway. No matter. What matters is that it's supersimple to define - even your kindergarten child could understand it - and yet it leads to strikingly beautiful, complex pictures that look like infinitely elaborate mosaics in mosques.
There is a board game called "chaos" that's based on sandpiles. The board is a grid, two players each have n grains of sand, they take turns placing a grain on the board, making little piles with the grains, and whenever a pile reaches height 4, it topples: each of the 4 grains gets moved to one of the 4 neighbors (in a predetermined order, in the case of the board game). Repeat toppling as needed. Any grain that falls off the board gets returned to its owner, and the winner is the player who first gets rid of all of their grains. That's the board game. Level: 8 years old and up.
In reality (by which I now mean, in the mathematical model), the grid is infinite, we start with n grains at the origin, keep toppling, and look at the final configuration as n goes to infinity, where every vertex holds a pile of grains of height 0,1, 2 or 3. The challenge is to analyze the convergence rate and the final configuration. Level: Math PhD and up.
Programming this might be a good first project: straightforward to implement, yet gives pretty pictures.
There is a board game called "chaos" that's based on sandpiles. The board is a grid, two players each have n grains of sand, they take turns placing a grain on the board, making little piles with the grains, and whenever a pile reaches height 4, it topples: each of the 4 grains gets moved to one of the 4 neighbors (in a predetermined order, in the case of the board game). Repeat toppling as needed. Any grain that falls off the board gets returned to its owner, and the winner is the player who first gets rid of all of their grains. That's the board game. Level: 8 years old and up.
In reality (by which I now mean, in the mathematical model), the grid is infinite, we start with n grains at the origin, keep toppling, and look at the final configuration as n goes to infinity, where every vertex holds a pile of grains of height 0,1, 2 or 3. The challenge is to analyze the convergence rate and the final configuration. Level: Math PhD and up.
Programming this might be a good first project: straightforward to implement, yet gives pretty pictures.
Tuesday, June 28, 2011
Hotels, and the marginal utility of wealth
I have stayed in three different hotels in the past seven days.
High-end hotel. Unctuous staff. Complimentary wine. Airport music in the lobby and elevator. Lemon-flavored water. High-tech light switches that require a college degree to use. Abstract art. Everything is made to stand out ostentatiously. Elongated shampoo bottle, zig-zaging floor pattern, and a bed big enough to get lost in. A good place to feel dizzy. No ironing board.
Comfortable business hotel (160ドル plus tax). Bland in the extreme. Staff blends in with the walls. No colors. Nothing stands out at all. Clean, anonymous, comfortable, it could be situated anywhere in the world. A good place for an out-of-body experience. Ironing board.
Hostel (17ドル.10 including breakfast). Small bouncy bed. Matter-of-fact staff. Friendly people gathering around the outdoor pool. Room-sharing with a stranger who turns out to be, of all things, an assistant professor of computer science who asked me what I worked on and who, when I mentioned bin packing, said it reminded her of knapsack problems that come up in cryptography!! A good place to make friends. No ironing board.
In my case the marginal utility of money for lodging is negative.
High-end hotel. Unctuous staff. Complimentary wine. Airport music in the lobby and elevator. Lemon-flavored water. High-tech light switches that require a college degree to use. Abstract art. Everything is made to stand out ostentatiously. Elongated shampoo bottle, zig-zaging floor pattern, and a bed big enough to get lost in. A good place to feel dizzy. No ironing board.
Comfortable business hotel (160ドル plus tax). Bland in the extreme. Staff blends in with the walls. No colors. Nothing stands out at all. Clean, anonymous, comfortable, it could be situated anywhere in the world. A good place for an out-of-body experience. Ironing board.
Hostel (17ドル.10 including breakfast). Small bouncy bed. Matter-of-fact staff. Friendly people gathering around the outdoor pool. Room-sharing with a stranger who turns out to be, of all things, an assistant professor of computer science who asked me what I worked on and who, when I mentioned bin packing, said it reminded her of knapsack problems that come up in cryptography!! A good place to make friends. No ironing board.
In my case the marginal utility of money for lodging is negative.
Monday, June 27, 2011
Program committees on the weekend: a question of priorities
Why do program committee meetings often take place during the weekend? Because people have other work commitments during the week, and it is hard to get them all together at the same time.
But what about family commitments? Those don't count. People have some flexibility and can always shift things around, it is assumed.
Why can't people shift things around at work as well? Because some of those are hard constraints.
But at home also, some constraints are hard: of a child is performing on a school play for example, that cannot be moved to another date.
I think that there is an unspoken assumption that gives priority to work over family time.
But what about family commitments? Those don't count. People have some flexibility and can always shift things around, it is assumed.
Why can't people shift things around at work as well? Because some of those are hard constraints.
But at home also, some constraints are hard: of a child is performing on a school play for example, that cannot be moved to another date.
I think that there is an unspoken assumption that gives priority to work over family time.
Sunday, June 26, 2011
Selling your honor for a job
How does one get an academic job in a competitive environment, surrounded by many equally qualified candidates? How can one have an edge, without actually changing one's research folder in any way? The book "Influence" suggests a number of marketing techniques that can presumably be applied to that setting, for example, while attending a conference just prior to hiring season.
One technique is called reciprocation. If someone does us a small favor, we are much more inclined to do them a favor later, even if it is completely out of proportion. So, at that conference, the eager job candidate can do all sorts of small favors for the members of hiring committees in attendance: call the elevator for them and hold the door; give them your umbrella if it's raining out; pay the tip at the restaurant; lend them your laptop, phone, connector, etc. Be on the lookout for small favors, and make sure to refuse to have them repaid to you in any way.
Another technique is called "commitment and consistency". If you can make them say something positive about your research, they will be much more likely later, when they review your folder, to think well of it, because they will want to be consistent with themselves. For example, ask them: "What would you say are my strong points?", or manage to extract some mention of a harmless commitment: say, if you happen to be in town or on campus for some other reason, could you come by at the same time? Or you could tell the person how much you appreciate the attention they pay to all candidates, how they are particularly careful to be fair towards this or that category of applicants to avoid discrimination, or otherwise clothe them with some quality of character that they might accept; then later, for consistency, the will feel obligated to act in that way.
Another technique is "social proof": if someone is unsure how to evaluate a situation or person, they will look at other people's reactions to form their own opinions. You could enroll some other student who could subtly intimate, by their attitude or words, that you are an awesome star. You could have "clappers" in the audience when you give your talk: plant some friends who will clap or ask questions that will make you look good.
Another technique is "liking": you can make yourself likeable by, for example, being good-looking, being taller (wear heels), or trying to be similar to the people you want to suck up to. Dress like them, talk like them, watch them attentively and discover common interests in golf or books or whatever. Discover some connections. Go to their talk, know their work, find connections to your interests.
Another technique is "Authority": get a trusted authority to say good words on your behalf. I'm not sure how that can be done independently of the quality of your research, but maybe someone with more imagination than me can come up with it.
Another technique is "scarcity": convey the impression that you are sought after. You could have a busy schedule. You could have some fellow student mention that the department of x is showing much interest in you. You could say that you have some upcoming informal visit to some prestigious place.
My impression is that those marketing techniques are used very little in academia. Certainly, when I recognize an attempt at one of those, I am disgusted and write off the person doing it. But I can see how the skillful manipulator may be able to use them to have an edge.
One technique is called reciprocation. If someone does us a small favor, we are much more inclined to do them a favor later, even if it is completely out of proportion. So, at that conference, the eager job candidate can do all sorts of small favors for the members of hiring committees in attendance: call the elevator for them and hold the door; give them your umbrella if it's raining out; pay the tip at the restaurant; lend them your laptop, phone, connector, etc. Be on the lookout for small favors, and make sure to refuse to have them repaid to you in any way.
Another technique is called "commitment and consistency". If you can make them say something positive about your research, they will be much more likely later, when they review your folder, to think well of it, because they will want to be consistent with themselves. For example, ask them: "What would you say are my strong points?", or manage to extract some mention of a harmless commitment: say, if you happen to be in town or on campus for some other reason, could you come by at the same time? Or you could tell the person how much you appreciate the attention they pay to all candidates, how they are particularly careful to be fair towards this or that category of applicants to avoid discrimination, or otherwise clothe them with some quality of character that they might accept; then later, for consistency, the will feel obligated to act in that way.
Another technique is "social proof": if someone is unsure how to evaluate a situation or person, they will look at other people's reactions to form their own opinions. You could enroll some other student who could subtly intimate, by their attitude or words, that you are an awesome star. You could have "clappers" in the audience when you give your talk: plant some friends who will clap or ask questions that will make you look good.
Another technique is "liking": you can make yourself likeable by, for example, being good-looking, being taller (wear heels), or trying to be similar to the people you want to suck up to. Dress like them, talk like them, watch them attentively and discover common interests in golf or books or whatever. Discover some connections. Go to their talk, know their work, find connections to your interests.
Another technique is "Authority": get a trusted authority to say good words on your behalf. I'm not sure how that can be done independently of the quality of your research, but maybe someone with more imagination than me can come up with it.
Another technique is "scarcity": convey the impression that you are sought after. You could have a busy schedule. You could have some fellow student mention that the department of x is showing much interest in you. You could say that you have some upcoming informal visit to some prestigious place.
My impression is that those marketing techniques are used very little in academia. Certainly, when I recognize an attempt at one of those, I am disgusted and write off the person doing it. But I can see how the skillful manipulator may be able to use them to have an edge.
Saturday, June 25, 2011
A primal-dual analysis of the Ranking algorithm
The other day I was sitting next to Kamal Jain on a flight, and he told me about a neat primal-dual analysis (due to Nikhil Devanur, Bobby Kleinberg, and him) of the Karp-Vazirani-Vazirani Ranking algorithm for online matching. Since the adwords result also has a primal-dual analysis, this is a step towards unifying the results.
The n vertices on one side of the bipartite graph (indexed by i) are known in advance, the n vertices on the other side (indexed by j) arrive online, and the edges adjacent to j are revealed upon j's arrival.
Here is one way to present the Ranking algorithm: each i picks a random value y(i) independently drawn from the uniform distribution over [0,1]. When j arrives, is at least one of its neighbors is not yet matched, j gets matched to the available neighbor with smallest value y(i). Every time an edge e is added to the matching the corresponding variable x(e) is set to 1 in the standard linear programming formulation for maximum bipartite matching.
Here is how the dual is defined. There is one variable b(i) for each i, which stays at 0 as long as i is not matched, and one variable a(j) for each j, which will be set upon j's arrival. We use a water-level algorithm, so initially every i is filled with water up to level 0, and the level only goes up. When j arrives, consider the available neighbors of j. In a continuous manner, take the ones filled to the lowest level, and raise their water level together (new available neighbors of j may join in along the way if the water level reaches their current water level.) This stops when some i reaches level y(i). At that time, (j is matched to i), we set variable b(i) to (exp(y(i))/(e-1), and variable a(j) to (e-exp(y(i))/(e-1). If j stays unmatched, we set a(j) to 0.
To relate the value of the primal (number of matching edges) to the value of the dual (sum over i of a(i) plus sum over j of a(j)), we consider two cases: if j is not matched, neither the primal nor the dual change values. If j is matched, the primal increases by 1 and the dual increases by (exp(y(i))/(e-1)+(e-exp(y(i))/(e-1), which equals 1/(1-1/e), as desired. This the value of the primal is at least (1-1/e) times the value of the dual.
Obviously, the primal is feasible: it's a matching.
Unfortunately the dual might be unfeasible: the dual constraints, a(j)+b(i)>= 1 for each edge {i,j}, might be violated. The idea is that the dual solution is feasible on average. Then the expected value of the output matching will be at least (1-1/e) times the expected value of the dual, which, by feasibility of the average, is greater than (1-1/e) times OPT.
To prove that the dual is feasible on average, focus on one edge {i,j}, fix the value of y(i') arbitrarily for all i' different from i, and consider the executions of the algorithms as y(i) ranges through all values in [0,1]. In addition, consider an artificial execution in which i is not present, and for that execution, let L denote the water level at which j is matched, if there is such a level; L=1 if j stays unmatched. For that artificial execution, a(j) equals (e-exp(L))/(e-1). Now, put i back into the system and continuously decrease y(i) from 1 to zero. By monotonicity of the Ranking algorithm, a decrease in y(i) cannot decrease a(j), so a(j) will be at least (e-exp(L))/(e-1) in all executions. On the other hand, as soon as y(i) goes below L, edge {i,j} appears in the matching, and by monotonicity, a decrease in y(i) cannot cause vertex i to go from being matched to being unmatched, so b(i)=exp(y(i))/(e-1) for all values of y(i) in [0,L]. Integrating over y(i), we obtain that the value of E(a(j))+E(b(i)) is at least 1, so constraint {i,j} is satisfied. This proves feasibility of the average of the duals.
The n vertices on one side of the bipartite graph (indexed by i) are known in advance, the n vertices on the other side (indexed by j) arrive online, and the edges adjacent to j are revealed upon j's arrival.
Here is one way to present the Ranking algorithm: each i picks a random value y(i) independently drawn from the uniform distribution over [0,1]. When j arrives, is at least one of its neighbors is not yet matched, j gets matched to the available neighbor with smallest value y(i). Every time an edge e is added to the matching the corresponding variable x(e) is set to 1 in the standard linear programming formulation for maximum bipartite matching.
Here is how the dual is defined. There is one variable b(i) for each i, which stays at 0 as long as i is not matched, and one variable a(j) for each j, which will be set upon j's arrival. We use a water-level algorithm, so initially every i is filled with water up to level 0, and the level only goes up. When j arrives, consider the available neighbors of j. In a continuous manner, take the ones filled to the lowest level, and raise their water level together (new available neighbors of j may join in along the way if the water level reaches their current water level.) This stops when some i reaches level y(i). At that time, (j is matched to i), we set variable b(i) to (exp(y(i))/(e-1), and variable a(j) to (e-exp(y(i))/(e-1). If j stays unmatched, we set a(j) to 0.
To relate the value of the primal (number of matching edges) to the value of the dual (sum over i of a(i) plus sum over j of a(j)), we consider two cases: if j is not matched, neither the primal nor the dual change values. If j is matched, the primal increases by 1 and the dual increases by (exp(y(i))/(e-1)+(e-exp(y(i))/(e-1), which equals 1/(1-1/e), as desired. This the value of the primal is at least (1-1/e) times the value of the dual.
Obviously, the primal is feasible: it's a matching.
Unfortunately the dual might be unfeasible: the dual constraints, a(j)+b(i)>= 1 for each edge {i,j}, might be violated. The idea is that the dual solution is feasible on average. Then the expected value of the output matching will be at least (1-1/e) times the expected value of the dual, which, by feasibility of the average, is greater than (1-1/e) times OPT.
To prove that the dual is feasible on average, focus on one edge {i,j}, fix the value of y(i') arbitrarily for all i' different from i, and consider the executions of the algorithms as y(i) ranges through all values in [0,1]. In addition, consider an artificial execution in which i is not present, and for that execution, let L denote the water level at which j is matched, if there is such a level; L=1 if j stays unmatched. For that artificial execution, a(j) equals (e-exp(L))/(e-1). Now, put i back into the system and continuously decrease y(i) from 1 to zero. By monotonicity of the Ranking algorithm, a decrease in y(i) cannot decrease a(j), so a(j) will be at least (e-exp(L))/(e-1) in all executions. On the other hand, as soon as y(i) goes below L, edge {i,j} appears in the matching, and by monotonicity, a decrease in y(i) cannot cause vertex i to go from being matched to being unmatched, so b(i)=exp(y(i))/(e-1) for all values of y(i) in [0,L]. Integrating over y(i), we obtain that the value of E(a(j))+E(b(i)) is at least 1, so constraint {i,j} is satisfied. This proves feasibility of the average of the duals.
Friday, June 24, 2011
Writing a proof is like translating
When you try to translate a text from one language to another, you are often faced with words that have no exact translation. The exact concept does not exist in the receptor language. So you have to make do with an approximate translation, either keeping the nuances that you judge to be more important and reluctantly dropping the others, or writing long, cumbersome sentences that cram in every notion you can think of, at the cost of the entire disappearance of the original elegance of style.
When you try to write a mathematical proof, it's a little bit of the same process. You have an idea in your head, an organized ensemble of mathematical notions, and you try to communicate them with symbols and English words, but what is written can never be exactly what was in your mind, and you have to make compromises. Sometimes you can spend hours, no, days, going back and forth: "I want to say this first. But then, that forces me to make a certain other choice. And then, there are further consequences. Oh well, at this point the changes are extensive and overall I am not sure I gained anything; I lost that other aspect that I really wanted to convey...", and it is really easy to go around in circles.
That's where our training in runtime and tradeoffs between efficiency and approximation ought to come into play! We can use hard constraints on runtime: "I will stop writing and send out what I have on Friday, regardless of how far it is from perfect"; or on quality: "As soon as the proof is written in full detail and hangs together with no gaps, I will put it on arxiv, even if it's unnecessarily cumbersome". Soft constraints are dangerous: "I will stop tinkering with the paper as soon as I see that an extra day of work does not lead to any significant improvements". When co-authors have different standards, it can lead to tensions.
The better written it is, the more readers there will be. In the limit, if it's perfectly well written, which takes infinite time, then everyone will read it.
When you try to write a mathematical proof, it's a little bit of the same process. You have an idea in your head, an organized ensemble of mathematical notions, and you try to communicate them with symbols and English words, but what is written can never be exactly what was in your mind, and you have to make compromises. Sometimes you can spend hours, no, days, going back and forth: "I want to say this first. But then, that forces me to make a certain other choice. And then, there are further consequences. Oh well, at this point the changes are extensive and overall I am not sure I gained anything; I lost that other aspect that I really wanted to convey...", and it is really easy to go around in circles.
That's where our training in runtime and tradeoffs between efficiency and approximation ought to come into play! We can use hard constraints on runtime: "I will stop writing and send out what I have on Friday, regardless of how far it is from perfect"; or on quality: "As soon as the proof is written in full detail and hangs together with no gaps, I will put it on arxiv, even if it's unnecessarily cumbersome". Soft constraints are dangerous: "I will stop tinkering with the paper as soon as I see that an extra day of work does not lead to any significant improvements". When co-authors have different standards, it can lead to tensions.
The better written it is, the more readers there will be. In the limit, if it's perfectly well written, which takes infinite time, then everyone will read it.
Thursday, June 23, 2011
Women, academia, tenure, children, life
Last week Valerie King and I, later joined by Cora Borradaile, had a conversation on skype about women, academic job choices, and tenure. Here is the transcript of our conversation.
[06/17/11] Claire Mathieu: Hi Val! Here's the quote you sent me from the Barnard Commencement speech.
Sheryl Sandberg
Chief operating officer, Facebook
Barnard College
Women almost never make one decision to leave the work force. It doesn't happen that way. They make small little decisions along the way that eventually lead them there. Maybe it's the last year of med school when they say, I'll take a slightly less interesting specialty because I'm going to want more balance one day. Maybe it's the fifth year in a law firm when they say, I'm not even sure I should go for partner, because I know I'm going to want kids eventually.These women don't even have relationships, and already they're finding balance, balance for responsibilities they don't yet have. And from that moment, they start quietly leaning back.
So, my heartfelt message to all of you is, and start thinking about this now, do not leave before you leave. Do not lean back; lean in. Put your foot on that gas pedal and keep it there until the day you have to make a decision, and then make a decision. That's the only way, when that day comes, you'll even have a decision to make.
[06/17/11] Valerie King: When I read the quote, I felt it described to some extent what I had done myself. That is, I thought, I should get a job at a university where I could raise a family and get tenure at the same time, without worrying too much about it. I decided this even before I was marrried!
[06/17/11] Claire Mathieu: Were you worried about not having time for a family if you went to a fiercely competitive place like, say, MIT?
[06/17/11] Valerie King: I don't think I ever had the option to take a prof position at MIT but I did have an offer to go to Brown for example. Way before I did the postdocs and NEC stint.
[06/17/11] Claire Mathieu: Did you really think that a postdoc was a safer choice than Brown?
[06/17/11] Valerie King: I was scared to take a permanent position somehow, being single. I think the family idea was always in the back of my mind.
[06/17/11] Claire Mathieu: It's scary for a man too.
[06/17/11] Valerie King: You mean the idea of settling in one place?
[06/17/11] Claire Mathieu: Yes, especially while still single.
[06/17/11] Valerie King: Hmm, maybe you're right about that. But I do know that I wanted to go to a place where I could be reasonably sure of getting tenure while raising kids. I wonder if that quote actually contains good advice? I mean, when I think about the women who have done well, a lot of them have established themselves at research labs first and then gone straight to tenured jobs.
[06/17/11] Claire Mathieu: You mean, Anna Karlin, Joan Feigenbaum,...?
[06/17/11] Valerie King: Yup, I'm trying to think of examples. Can you imagine someone going to a high pressure place, trying to juggle children and tenure, getting denied tenure, and then having to look, with little kids?? When I had little kids, I would get maybe 4 hours of sleep, get up at 5:00 in the morning, start class preparation. I was a walking zombie, not really able to plan things, or do research much.
[[06/17/11] Claire Mathieu: No, it's very stressful. I was so lucky that in France my job was tenured right away. I never had to face those worries. Also, I could take it easy when my children were very little. I joined CNRS, that had no teaching, in 1990, and stayed until 1997, and my children were born in 1991 and 1994. During those years, if they had a bad night's sleep, for example, I could sleep late in the morning. I taught a little bit every year, just for fun, but not much. My parents helped take care of the children when I traveled to conferences, and I could spend several months each year in the US (taking my children with me) visiting other departments and building research connections. The pay was minimal but it was an ideal position to get started in research, and the stress was inexistent. So that quote doesn't really resonate with me. I never had to make, or even face the possibility of having to make, hard choices between children and career.
[06/17/11] Valerie King: Wow. So if one was to follow the advice given in that quote, one would just go off to a high stress place, and then when and if one had kids, and felt the pressure, one would just try to find another job, I guess.
[06/17/11] Claire Mathieu: That requires a lot of trust that things will work out somehow. Maybe it's ok if there are things that you know you enjoy and are confident that you could always do. Some fallback plan. For example if you know you'd enjoy being a high school teacher, that's a fallback plan if your research career ambitions don't work out. I always knew I liked teaching, and that I could be good at teacher if I put my heart into it, so that was always my plan B...
06/17/11] Valerie King: I see. Go for the high risk position. Then fall back if it's necessary.
[06/17/11] Claire Mathieu: You have a law degree: aren't there other things you would have done if academia didn't work out for you?
[06/17/11] Valerie King: It's hard to know. I certainly didn't think that way. Because law is not that different from academia. If I had stayed so many years away from the law, who would have hired me? And in Canada, being a US lawyer? It's also a competitive occupation.
[06/17/11] Claire Mathieu: I have a friend, an academic, who had some interest in early childhood education and could imagine working in a kindergarten.
[06/17/11] Valerie King: Have you heard of many academics around our age who are considering changing careers?
[06/17/11] Claire Mathieu: Um... you mean, not becoming dean or provost or university president and all that stuff? Maria Klawe. David Dobkin.
[06/17/11] Valerie King: Right. There's the admin route.
[06/17/11] Claire Mathieu: Right. Oh, horror!
[06/17/11] Valerie King: Horror for you?
[06/17/11] Claire Mathieu: (Yes). I have a neighbor, Richard Schwartz, a mathematician, who wrote a book for children that was bestseller on Amazon. Maybe he could switch and become a writer!
[06/17/11] Valerie King: Wow. I have a friend who is considering piecing together an alternate career made up of tutoring, teaching yoga, consulting.
[06/17/11] Claire Mathieu: Why?
[06/17/11] Valerie King: I'm not sure. Too much stress. For me, it would be boredom of doing the same thing (teaching) in the same place with the same people year after year (20! years)
[06/17/11] Claire Mathieu: I admire Jon Kleinberg. The way in which he just moved forward in the direction he was interested in.
[06/17/11] Valerie King: What Kleinberg does though is not necessarily theoretical computer science. To some extent it's sociology
[06/17/11] Claire Mathieu: I agree, it's not necessarily TCS, but does that matter? Jon Kleinberg was not thinking about how other academics would see him, I think. He's just doing what interests him, whatever happens. It's daring. He wasn't advocating about how people ought to do research in this or that: he just does it. His attitude is great.
[06/17/11] Valerie King: I think Les Valiant is the same, and great for the same reason.
[06/17/11] Claire Mathieu: Cora Borradaile is online. Should I invite her to join in?
[06/17/11] Valerie King: Sure, is she interested?
[06/17/11] Glencora Borradaile: Hello?
[06/17/11] Valerie King: Hi Cora!
[06/17/11] Claire Mathieu: Hi Cora, welcome!
[06/17/11] Glencora Borradaile: Hi!
[06/17/11] Claire Mathieu: What's the stress when you have tenure? John Savage told me that if what I do is not satisfying, then I can always find other ways to find fulfillment in academia. For example, I could think about writing a book (technical or broader), or focus on advising first-year undergrads.
[06/17/11] Valerie King: I find that when I'm teaching two courses at the same time, I'm kind of unable to do much extra. I seem to get caught up in the daily grind with teaching, admin etc. maybe since I still have my kids at home. I don't seem to have that much time for changing direction.
[06/17/11] Claire Mathieu: Oh I see. You can teach a different course, or sit on a different committee, or go to a different conference, but that's about the extent of your flexibility. Well, what would you like to do? If you had the time? Professionally, I mean.
06/17/11] Valerie King: I don't know. Maybe get out of the university.
[06/17/11] Glencora Borradaile: When I hear these reasons for women "opting out" or "opting for a less high-profile position", it is almost always linked to kids, but there is something there even if kids aren't in the picture
[06/17/11] Valerie King: I'm reaching the age where the kids will be out of the house, so I'm contemplating what to do next. I suppose I would like to think and work on problems in a broader context. I'd love to be in a think tank.
[06/17/11] Glencora Borradaile: i've started getting involved in Corvallis - that's been really satisfying - to help things happen in the city
[06/17/11] Valerie King: I've thought about getting involved with local planning in Victoria. What are you doing in Corvallis?
[06/17/11] Glencora Borradaile: Mostly it is boring meetings, but there is actually interesting issues and i can't help but think of mathematical solutions.
[06/17/11] Claire Mathieu: Like the stuff Pascal van Hentenryck does, constraint programming in real life?
[06/17/11] Glencora Borradaile: More like route design, Claire. I'm involved in bicycle and pedestrian advocacy. There's room for really changing the structure of the city by putting in bike paths and multi-use paths. If you could drop all the traffic lights and one-way streets, how would you make traffic move? if you have X dollars, what is the best bicycle boulevard you could build? [06/17/11] Claire Mathieu: Perfect for you!
[06/17/11] Glencora Borradaile: Yes! Unfortunately it's not a clean problem - too many rules and land-use issues
[06/17/11] Valerie King: I do think this is related to having children or not. With kids at home, I felt obligated to spend time with them in the evenings. Probably it was not a good idea to do this exclusively. The mother role together with the academic role was fine, totally consuming. I couldn't have imagined doing extra work. Though now I have time. And I did start to do some volunteer work around 6 years ago.
[06/17/11] Glencora Borradaile: So are you talking about replacing the outside-academia work, or replacing the academia work?
[06/17/11] Claire Mathieu: I was thinking academia. You're right, let's stay focused.
[06/17/11] Glencora Borradaile: I might have not absorbed the whole of the earlier conversation - why are you dissatisfied with academia that makes you want to replace it?
[06/17/11] Valerie King: For me, it's because I'm bored with doing it for the past 20 years.
[06/17/11] Claire Mathieu: Me too but I won't say it
[06/17/11] Valerie King: You just did
[06/17/11] Claire Mathieu: No I didn't :)
[06/17/11] Valerie King: Why won't you say it?
[06/17/11] Claire Mathieu: Well, I don't know. I'm not THAT bored. When I start discussing a problem with someone, it's always fun
[06/17/11] Valerie King: So the funny thing is that tenure doesn't change anything that much. People say you're like a trained rat.
[06/17/11] Glencora Borradaile: Dan Spielman told me it's worse when you have tenure
[06/17/11] Valerie King: Really? why?
[06/17/11] Glencora Borradaile: Because of the committee work and added letter writing (and there may have been other things that I don't remember)
[06/17/11] Claire Mathieu: He sounds like a very responsible person
[06/17/11] Valerie King: Oh there's a ton of stuff that fills up your time
[06/17/11] Glencora Borradaile: Heh
[06/17/11] Claire Mathieu: The thing is, if you start slacking, once you have tenure, nothing bad happens to you so it's your choice just how much time you spend on things; besides, if you give all your energy to work, how can they expect more?
[06/17/11] Glencora Borradaile: Maybe Dan can't say no.
[06/17/11] Claire Mathieu: And then he feels obligated to do it
[06/17/11] Valerie King: I think the committee work is really up to the individual. Some people work hard on it, some don't. But I do feel obligated to write letters, help out my students, do some refereeing, etc.
[06/17/11] Claire Mathieu: Writing letters is a big and important chore. Helping students is fun for me. Refereeing... well... maybe I've done enough at this point. Better say no right away than let it drag for months because I'm not really into it, right?
[06/17/11] Valerie King: Do you think we can stop refereeing now, with a clear conscience?
[06/17/11] Glencora Borradaile: Noooo! that would just mean more for the rest of us!
[06/17/11] Claire Mathieu: What do you mean, conscience? If you spend your time doing some other work, it's just replacing work by work. I hope that if we do more of the things we enjoy and less of the things we can't bear any more, our job will be more fun
[06/17/11] Valerie King: Really? But I thought no one wants to do this, so if we don't pitch in, it won't get done
[06/17/11] Valerie King: Maybe I'll stop refereeing
[06/17/11] Glencora Borradaile: i've started counting ... for every paper I submit, I will review 3 papers and feel perfectly happy saying no to any requests after my quota is met
[06/17/11] Valerie King: What about program committees? Do you see any point of being on one of these?
[06/17/11] Claire Mathieu: I used to find that task really interesting: I've past that stage, I think, so I may say no from now on.
[06/17/11] Valerie King: Every year, I agree to 1-3 or so. I don't like it either.
[06/17/11] Claire Mathieu: I mean, why should you do anything? If boredom prevents you from doing your work, shouldn't you re-evaluate? Maybe eliminate the worst chores and emphasize others?
[06/17/11] Valerie King: I don't know where this comes from. A desire to prove myself still?
[06/17/11] Claire Mathieu: Val, I think that that's John Savage's point: you get to define the job. If you don't like it, re-define it.
[06/17/11] Glencora Borradaile: You agree to 1-3 PC's EVERY year, Valerie?
[06/17/11] Valerie King: Yes. You know, in my department, in other areas, people are doing a lot more than this. SODA, FOCS and STOC take a lot of time. Others take half as much or less.
06/17/11] Valerie King: Should we tie it up? I have to make dinner.
[06/17/11] Claire Mathieu: Ok, what do we conclude?
[06/17/11] Valerie King: We conclude that we're not sure about the quote. It seems it's very risky to do what she's proposing in academia.
[06/17/11] Claire Mathieu: Unless you have a plan B.
[06/17/11] Valerie King: And we need to take steps to keep up our interest. Change jobs, change the tasks we do, give up the things we don't like to do.
[06/17/11] Glencora Borradaile: i have a problem with any advice that advocates work before all else - in the absence of the kid question
[06/17/11] Valerie King: I only get creative when the work is really on my mind, at least temporarily. This does interfere with kids though
[06/17/11] Claire Mathieu: Work before all else? Did we advocate that? Even the quote, I think, advocated work before "the potential unknown". not work before something else that would be well-defined.
[06/17/11] Valerie King: Right, Claire.
[06/17/11] Glencora Borradaile: "Put your foot on that gas pedal and keep it there" isn't a healthy phrase for me
[06/17/11] Claire Mathieu: Biker
[06/17/11] Glencora Borradaile: Haha
[06/17/11] Valerie King: It's hard to imagine research done that way. I have to take naps every few hours when I'm really working. But I am pretty one track minded. Maybe it's a question of personality.
[06/17/11] Claire Mathieu: When I'm "really working", the social part of my life, limited as it is, takes a toll. I find find harder and harder to accept, with time. The thrill of understanding a mathematical structure, compared to the steady pleasure of seeing my kids or talking to regular people... the balance is changing. I can imagine stopping to do research at some point if I find something else that I love, even if it's not as much of an intellectual stimulation.
[06/17/11] Glencora Borradaile: Something i've wondered - if you do "lean in" would you even recognize that your life could be different and you want to have kids or join a motorcycle gang or whatever it is? Would you be able to make such a change?
[06/17/11] Claire Mathieu: Huh?
[06/17/11] Valerie King: And I was thinking I would do more research when the kids left. I don't know yet. I was also thinking of doing more sociological type research.
[06/17/11] Valerie King: Better end this.
[06/17/11] Claire Mathieu: Ok. Bye?
[06/17/11] Glencora Borradaile: i have to run (not literally)
[06/17/11] Claire Mathieu: Ok, bye.
[06/17/11] Claire Mathieu: Hi Val! Here's the quote you sent me from the Barnard Commencement speech.
Sheryl Sandberg
Chief operating officer, Facebook
Barnard College
Women almost never make one decision to leave the work force. It doesn't happen that way. They make small little decisions along the way that eventually lead them there. Maybe it's the last year of med school when they say, I'll take a slightly less interesting specialty because I'm going to want more balance one day. Maybe it's the fifth year in a law firm when they say, I'm not even sure I should go for partner, because I know I'm going to want kids eventually.These women don't even have relationships, and already they're finding balance, balance for responsibilities they don't yet have. And from that moment, they start quietly leaning back.
So, my heartfelt message to all of you is, and start thinking about this now, do not leave before you leave. Do not lean back; lean in. Put your foot on that gas pedal and keep it there until the day you have to make a decision, and then make a decision. That's the only way, when that day comes, you'll even have a decision to make.
[06/17/11] Valerie King: When I read the quote, I felt it described to some extent what I had done myself. That is, I thought, I should get a job at a university where I could raise a family and get tenure at the same time, without worrying too much about it. I decided this even before I was marrried!
[06/17/11] Claire Mathieu: Were you worried about not having time for a family if you went to a fiercely competitive place like, say, MIT?
[06/17/11] Valerie King: I don't think I ever had the option to take a prof position at MIT but I did have an offer to go to Brown for example. Way before I did the postdocs and NEC stint.
[06/17/11] Claire Mathieu: Did you really think that a postdoc was a safer choice than Brown?
[06/17/11] Valerie King: I was scared to take a permanent position somehow, being single. I think the family idea was always in the back of my mind.
[06/17/11] Claire Mathieu: It's scary for a man too.
[06/17/11] Valerie King: You mean the idea of settling in one place?
[06/17/11] Claire Mathieu: Yes, especially while still single.
[06/17/11] Valerie King: Hmm, maybe you're right about that. But I do know that I wanted to go to a place where I could be reasonably sure of getting tenure while raising kids. I wonder if that quote actually contains good advice? I mean, when I think about the women who have done well, a lot of them have established themselves at research labs first and then gone straight to tenured jobs.
[06/17/11] Claire Mathieu: You mean, Anna Karlin, Joan Feigenbaum,...?
[06/17/11] Valerie King: Yup, I'm trying to think of examples. Can you imagine someone going to a high pressure place, trying to juggle children and tenure, getting denied tenure, and then having to look, with little kids?? When I had little kids, I would get maybe 4 hours of sleep, get up at 5:00 in the morning, start class preparation. I was a walking zombie, not really able to plan things, or do research much.
[[06/17/11] Claire Mathieu: No, it's very stressful. I was so lucky that in France my job was tenured right away. I never had to face those worries. Also, I could take it easy when my children were very little. I joined CNRS, that had no teaching, in 1990, and stayed until 1997, and my children were born in 1991 and 1994. During those years, if they had a bad night's sleep, for example, I could sleep late in the morning. I taught a little bit every year, just for fun, but not much. My parents helped take care of the children when I traveled to conferences, and I could spend several months each year in the US (taking my children with me) visiting other departments and building research connections. The pay was minimal but it was an ideal position to get started in research, and the stress was inexistent. So that quote doesn't really resonate with me. I never had to make, or even face the possibility of having to make, hard choices between children and career.
[06/17/11] Valerie King: Wow. So if one was to follow the advice given in that quote, one would just go off to a high stress place, and then when and if one had kids, and felt the pressure, one would just try to find another job, I guess.
[06/17/11] Claire Mathieu: That requires a lot of trust that things will work out somehow. Maybe it's ok if there are things that you know you enjoy and are confident that you could always do. Some fallback plan. For example if you know you'd enjoy being a high school teacher, that's a fallback plan if your research career ambitions don't work out. I always knew I liked teaching, and that I could be good at teacher if I put my heart into it, so that was always my plan B...
06/17/11] Valerie King: I see. Go for the high risk position. Then fall back if it's necessary.
[06/17/11] Claire Mathieu: You have a law degree: aren't there other things you would have done if academia didn't work out for you?
[06/17/11] Valerie King: It's hard to know. I certainly didn't think that way. Because law is not that different from academia. If I had stayed so many years away from the law, who would have hired me? And in Canada, being a US lawyer? It's also a competitive occupation.
[06/17/11] Claire Mathieu: I have a friend, an academic, who had some interest in early childhood education and could imagine working in a kindergarten.
[06/17/11] Valerie King: Have you heard of many academics around our age who are considering changing careers?
[06/17/11] Claire Mathieu: Um... you mean, not becoming dean or provost or university president and all that stuff? Maria Klawe. David Dobkin.
[06/17/11] Valerie King: Right. There's the admin route.
[06/17/11] Claire Mathieu: Right. Oh, horror!
[06/17/11] Valerie King: Horror for you?
[06/17/11] Claire Mathieu: (Yes). I have a neighbor, Richard Schwartz, a mathematician, who wrote a book for children that was bestseller on Amazon. Maybe he could switch and become a writer!
[06/17/11] Valerie King: Wow. I have a friend who is considering piecing together an alternate career made up of tutoring, teaching yoga, consulting.
[06/17/11] Claire Mathieu: Why?
[06/17/11] Valerie King: I'm not sure. Too much stress. For me, it would be boredom of doing the same thing (teaching) in the same place with the same people year after year (20! years)
[06/17/11] Claire Mathieu: I admire Jon Kleinberg. The way in which he just moved forward in the direction he was interested in.
[06/17/11] Valerie King: What Kleinberg does though is not necessarily theoretical computer science. To some extent it's sociology
[06/17/11] Claire Mathieu: I agree, it's not necessarily TCS, but does that matter? Jon Kleinberg was not thinking about how other academics would see him, I think. He's just doing what interests him, whatever happens. It's daring. He wasn't advocating about how people ought to do research in this or that: he just does it. His attitude is great.
[06/17/11] Valerie King: I think Les Valiant is the same, and great for the same reason.
[06/17/11] Claire Mathieu: Cora Borradaile is online. Should I invite her to join in?
[06/17/11] Valerie King: Sure, is she interested?
[06/17/11] Glencora Borradaile: Hello?
[06/17/11] Valerie King: Hi Cora!
[06/17/11] Claire Mathieu: Hi Cora, welcome!
[06/17/11] Glencora Borradaile: Hi!
[06/17/11] Claire Mathieu: What's the stress when you have tenure? John Savage told me that if what I do is not satisfying, then I can always find other ways to find fulfillment in academia. For example, I could think about writing a book (technical or broader), or focus on advising first-year undergrads.
[06/17/11] Valerie King: I find that when I'm teaching two courses at the same time, I'm kind of unable to do much extra. I seem to get caught up in the daily grind with teaching, admin etc. maybe since I still have my kids at home. I don't seem to have that much time for changing direction.
[06/17/11] Claire Mathieu: Oh I see. You can teach a different course, or sit on a different committee, or go to a different conference, but that's about the extent of your flexibility. Well, what would you like to do? If you had the time? Professionally, I mean.
06/17/11] Valerie King: I don't know. Maybe get out of the university.
[06/17/11] Glencora Borradaile: When I hear these reasons for women "opting out" or "opting for a less high-profile position", it is almost always linked to kids, but there is something there even if kids aren't in the picture
[06/17/11] Valerie King: I'm reaching the age where the kids will be out of the house, so I'm contemplating what to do next. I suppose I would like to think and work on problems in a broader context. I'd love to be in a think tank.
[06/17/11] Glencora Borradaile: i've started getting involved in Corvallis - that's been really satisfying - to help things happen in the city
[06/17/11] Valerie King: I've thought about getting involved with local planning in Victoria. What are you doing in Corvallis?
[06/17/11] Glencora Borradaile: Mostly it is boring meetings, but there is actually interesting issues and i can't help but think of mathematical solutions.
[06/17/11] Claire Mathieu: Like the stuff Pascal van Hentenryck does, constraint programming in real life?
[06/17/11] Glencora Borradaile: More like route design, Claire. I'm involved in bicycle and pedestrian advocacy. There's room for really changing the structure of the city by putting in bike paths and multi-use paths. If you could drop all the traffic lights and one-way streets, how would you make traffic move? if you have X dollars, what is the best bicycle boulevard you could build? [06/17/11] Claire Mathieu: Perfect for you!
[06/17/11] Glencora Borradaile: Yes! Unfortunately it's not a clean problem - too many rules and land-use issues
[06/17/11] Valerie King: I do think this is related to having children or not. With kids at home, I felt obligated to spend time with them in the evenings. Probably it was not a good idea to do this exclusively. The mother role together with the academic role was fine, totally consuming. I couldn't have imagined doing extra work. Though now I have time. And I did start to do some volunteer work around 6 years ago.
[06/17/11] Glencora Borradaile: So are you talking about replacing the outside-academia work, or replacing the academia work?
[06/17/11] Claire Mathieu: I was thinking academia. You're right, let's stay focused.
[06/17/11] Glencora Borradaile: I might have not absorbed the whole of the earlier conversation - why are you dissatisfied with academia that makes you want to replace it?
[06/17/11] Valerie King: For me, it's because I'm bored with doing it for the past 20 years.
[06/17/11] Claire Mathieu: Me too but I won't say it
[06/17/11] Valerie King: You just did
[06/17/11] Claire Mathieu: No I didn't :)
[06/17/11] Valerie King: Why won't you say it?
[06/17/11] Claire Mathieu: Well, I don't know. I'm not THAT bored. When I start discussing a problem with someone, it's always fun
[06/17/11] Valerie King: So the funny thing is that tenure doesn't change anything that much. People say you're like a trained rat.
[06/17/11] Glencora Borradaile: Dan Spielman told me it's worse when you have tenure
[06/17/11] Valerie King: Really? why?
[06/17/11] Glencora Borradaile: Because of the committee work and added letter writing (and there may have been other things that I don't remember)
[06/17/11] Claire Mathieu: He sounds like a very responsible person
[06/17/11] Valerie King: Oh there's a ton of stuff that fills up your time
[06/17/11] Glencora Borradaile: Heh
[06/17/11] Claire Mathieu: The thing is, if you start slacking, once you have tenure, nothing bad happens to you so it's your choice just how much time you spend on things; besides, if you give all your energy to work, how can they expect more?
[06/17/11] Glencora Borradaile: Maybe Dan can't say no.
[06/17/11] Claire Mathieu: And then he feels obligated to do it
[06/17/11] Valerie King: I think the committee work is really up to the individual. Some people work hard on it, some don't. But I do feel obligated to write letters, help out my students, do some refereeing, etc.
[06/17/11] Claire Mathieu: Writing letters is a big and important chore. Helping students is fun for me. Refereeing... well... maybe I've done enough at this point. Better say no right away than let it drag for months because I'm not really into it, right?
[06/17/11] Valerie King: Do you think we can stop refereeing now, with a clear conscience?
[06/17/11] Glencora Borradaile: Noooo! that would just mean more for the rest of us!
[06/17/11] Claire Mathieu: What do you mean, conscience? If you spend your time doing some other work, it's just replacing work by work. I hope that if we do more of the things we enjoy and less of the things we can't bear any more, our job will be more fun
[06/17/11] Valerie King: Really? But I thought no one wants to do this, so if we don't pitch in, it won't get done
[06/17/11] Valerie King: Maybe I'll stop refereeing
[06/17/11] Glencora Borradaile: i've started counting ... for every paper I submit, I will review 3 papers and feel perfectly happy saying no to any requests after my quota is met
[06/17/11] Valerie King: What about program committees? Do you see any point of being on one of these?
[06/17/11] Claire Mathieu: I used to find that task really interesting: I've past that stage, I think, so I may say no from now on.
[06/17/11] Valerie King: Every year, I agree to 1-3 or so. I don't like it either.
[06/17/11] Claire Mathieu: I mean, why should you do anything? If boredom prevents you from doing your work, shouldn't you re-evaluate? Maybe eliminate the worst chores and emphasize others?
[06/17/11] Valerie King: I don't know where this comes from. A desire to prove myself still?
[06/17/11] Claire Mathieu: Val, I think that that's John Savage's point: you get to define the job. If you don't like it, re-define it.
[06/17/11] Glencora Borradaile: You agree to 1-3 PC's EVERY year, Valerie?
[06/17/11] Valerie King: Yes. You know, in my department, in other areas, people are doing a lot more than this. SODA, FOCS and STOC take a lot of time. Others take half as much or less.
06/17/11] Valerie King: Should we tie it up? I have to make dinner.
[06/17/11] Claire Mathieu: Ok, what do we conclude?
[06/17/11] Valerie King: We conclude that we're not sure about the quote. It seems it's very risky to do what she's proposing in academia.
[06/17/11] Claire Mathieu: Unless you have a plan B.
[06/17/11] Valerie King: And we need to take steps to keep up our interest. Change jobs, change the tasks we do, give up the things we don't like to do.
[06/17/11] Glencora Borradaile: i have a problem with any advice that advocates work before all else - in the absence of the kid question
[06/17/11] Valerie King: I only get creative when the work is really on my mind, at least temporarily. This does interfere with kids though
[06/17/11] Claire Mathieu: Work before all else? Did we advocate that? Even the quote, I think, advocated work before "the potential unknown". not work before something else that would be well-defined.
[06/17/11] Valerie King: Right, Claire.
[06/17/11] Glencora Borradaile: "Put your foot on that gas pedal and keep it there" isn't a healthy phrase for me
[06/17/11] Claire Mathieu: Biker
[06/17/11] Glencora Borradaile: Haha
[06/17/11] Valerie King: It's hard to imagine research done that way. I have to take naps every few hours when I'm really working. But I am pretty one track minded. Maybe it's a question of personality.
[06/17/11] Claire Mathieu: When I'm "really working", the social part of my life, limited as it is, takes a toll. I find find harder and harder to accept, with time. The thrill of understanding a mathematical structure, compared to the steady pleasure of seeing my kids or talking to regular people... the balance is changing. I can imagine stopping to do research at some point if I find something else that I love, even if it's not as much of an intellectual stimulation.
[06/17/11] Glencora Borradaile: Something i've wondered - if you do "lean in" would you even recognize that your life could be different and you want to have kids or join a motorcycle gang or whatever it is? Would you be able to make such a change?
[06/17/11] Claire Mathieu: Huh?
[06/17/11] Valerie King: And I was thinking I would do more research when the kids left. I don't know yet. I was also thinking of doing more sociological type research.
[06/17/11] Valerie King: Better end this.
[06/17/11] Claire Mathieu: Ok. Bye?
[06/17/11] Glencora Borradaile: i have to run (not literally)
[06/17/11] Claire Mathieu: Ok, bye.
Wednesday, June 22, 2011
Marketing a paper: flattery
The book "Influence" by Robert Cialdini, that I read recently, has a short section on flattery, explaining that it is an effective way to get people to buy your product. This has been known forever, of course, but apparently it is effective even when the flattery is blatant.
How to apply that to marketing a paper: we know that the paper will most likely be reviewed by people who have done previous work in the same area. Thus, instead of presenting the results by contrasting them to previous results, with the implied message "Previous work was faulty, my work is much better", the writer should find a style that gives generous credit to previous work. For example, give ample credit to whoever "introduced the problem", "raised the question", "took the first steps", and, for intermediate results, "enhanced our understanding" or "made significant progress". Perhaps that it why we are told to cite the authors by name instead of merely referring to the article by number.
How to apply that to marketing a paper: we know that the paper will most likely be reviewed by people who have done previous work in the same area. Thus, instead of presenting the results by contrasting them to previous results, with the implied message "Previous work was faulty, my work is much better", the writer should find a style that gives generous credit to previous work. For example, give ample credit to whoever "introduced the problem", "raised the question", "took the first steps", and, for intermediate results, "enhanced our understanding" or "made significant progress". Perhaps that it why we are told to cite the authors by name instead of merely referring to the article by number.
Subscribe to:
Comments (Atom)