Dear Chair of Search Committee,
I have been asked by my current postdoctoral associate Joan Doe to write a letter of recommendation on her behalf, as she has expressed to me her interest for the tenure-track faculty position in MyAreaOfPhysics at Your Prestigious Institution. It is a pleasure for me to comply, as in my view Joan is a gifted young scientist, ready to take on the challenge of such a position. I have no doubt that she will succeed, that she will make an energetic colleague, and a fine addition to your already distinguished faculty.
It is my intent to make the strongest and most credible case possible in support of Joan’s application. Hence, in the remainder of this letter I shall do my best to focus on what I deem most relevant, while at the same time avoiding to write things that annoy, confuse or put me off, when I go through letters of recommendation written by others. Please note that, even though I generally write letters for physics applicants, it is my opinion that some general issues with letter writing are valid in any field, even outside the sciences.
Let me immediately make it clear to you that you are not receiving the same “boiler plate” letter that I have sent out to a dozen other places, some of them very different in character from yours. To the best of my ability, and to the extent that it is possible, I shall attempt to build a case for why, in my opinion, Joan makes a particularly suitable candidate for the position that Your Reputable Institution is seeking to fill, rather than for a generic faculty appointment at one of the many physics departments in North America. I am obviously genuinely convinced of Joan’s value as a solid scientist and researcher, as well as hard worker and effective communicator, thus capable of operating successfully in different settings; however, in this letter I am going to emphasize those aspects which I expect to be mostly relevant to Your department. This is why I mention Your Reputable Institution by name, spell out the position in my opening paragraph, and in the following try to tailor my letter so as to make Joan appear as a good candidate for you.
Because your institution is a research university, and because the position advertised is a standard tenure-track appointment, I am going to devote the bulk of my letter to discuss the candidate’s research ability. I shall also attempt to assess her teaching potential; clearly, however, just like for the vast majority of postdocs, I do not have as much evidence to offer a reliable judgment on this aspect.
Before my time
The first thing that I have to do, is give you an idea of how well I know the person . I have known Joan for two and a half years. She joined my group as a postdoctoral associate after receiving her PhD degree at MiddleOfThePack University , working under the supervision of Justone O’themany, one of the leading investigators in that subject area of which I am no expert. Her PhD thesis was on Really Hot Topic, one of great current interest in our field of science but one with which I am not directly familiar. I shall therefore avoid making a fool of myself trying to write about stuff I do not know, humbly deferring instead to her doctoral supervisor the task of assessing the significance of those accomplishments. Still, judging from her publication and citation records, it is clear to me that her doctoral research work has been very favourably received by the community. In particular, her 2007 article “Cool Findings on Really Hot Topic”, by herself and O’themany, has already garnered 43 citations, and Joan was invited in 2008 to give a talk on the subject at International Prominent Conference, hardly a common occurrence for a freshly minted PhD.
How did the candidate and I connect ?
When I received her inquiry about a possible postdoctoral opening in my group, I was surprised and a bit hesitant, as her past research work scarcely overlaps with my current interests. Given the pressure on junior scientists to accomplish something significant during their relatively short postdoctoral appointment, I was unsure whether joining my group would be a wise decision, as it would mean for her learning different techniques and becoming familiar with a whole new field. However, after reading the enthusiastic letter of support from O’themany, I decided to arrange for a meeting with her at the Annual Big Meeting. On that occasion, she stated to me her interest in broadening her research experience and background, as well as her desire to learn the techniques which I utilize, for she was thinking of applying them to tackle a Well-known Unsolved Problem (something of which I myself had frankly never thought). I was so impressed with her that I decided to make her an offer. This turned out to be an excellent decision. I have greatly enjoyed my interaction with Joan, who has not only proven a valuable asset to my research effort, she has also demonstrated independence and originality as researcher .
The candidate and me
The above is an introduction to the core of my letter, which consists of describing Joan’s research achievements during her time in my group. Obviously, this is the part about which I can speak most knowledgeably and authoritatively. It is also the part about which I am expected to write the most, at least when the position is one at a research university. Downplaying this part, emphasizing instead things like a candidate’s communicative skills, teaching accomplishments or potential, extra-curricular activities and so on (valuable as all that is), will have almost always the effect of generating the suspicion that the candidate does not possess above average research ability.
Now, the one thing I am not supposed to do here, is take this as an opportunity to tout my own horn, willingly or not taking the spotlight off Joan. I have read letters of recommendation in which the references took swipes at program directors for cutting their grants, or editors for rejecting their papers, or conference organizers for not inviting them. None of that belongs in here. This letter is about Joan, not me. The more I say about me, or anyone else, the more the reader will wonder why I am not writing more about Joan.
Along the same line, I am taking great care not to give the impression that Joan’s accomplishments over the past two years are merely the inevitable outcome of her being working with me, i.e., that under my sapient guidance, with the brilliant research project that I suggested, and the aid of the powerful methodology that I invented, anyone would have done well. The result of me doing that, would be for the reader to conclude that there is really nothing special about Joan, that she was just lucky enough to be working with me and she simply followed orders .
Instead, I am highlighting Joan’s individual contribution to the positive outcome of the project(s) in which she was engaged. This is the place for me to illustrate competently and clearly the importance and significance of the research carried out, as well as describe her key piece(s) of insight, her original idea(s), anything that should in my view be directly attributed to her.
I am making it absolutely clear that without Joan’s crucial collaboration, my group would not have arrived at the results that have been achieved (or, are within reach).
What are their strengths ?
At this point I really have to share with you what I have learned about Joan over the past few years, having enjoyed the opportunity of seeing her “in action”. These are the things that I am going to emphasize:
1) Creativity. Joan will be on her own, as an assistant professor. No one will be telling her how to write a grant proposal, suggesting projects for her and her students, or even how to teach her courses. A tenure-track assistant professor needs to be very resourceful, and for that imagination and creativity are required.
2) Technical ability. Imagination and creativity without technical ability amount to nothing . Simple as that. A good scientist must be able to do stuff, not just talk about it. She must be able carry out the work on her own (or teach students how to do it). The notion that “one can always collaborate with others” is a dangerous fallacy. The research program of a tenure-track faculty ought not hinge on the participation of others, especially supplying crucial technical support.
3) Productivity. At the cost of sounding prosaic, a good scientist has to be productive. That means, among other things, being able to get a project going and bringing it to fruition in a reasonable time span, as well as writing a good first draft of an article on a time scale of days or weeks, not months. Again, a tenure-track faculty will be expected to deliver on her own.
In making the above three points, I am trying to be as quantitative and precise as possible. When speaking of creativity, I describe one impressive idea; if I talk about technical ability I refer to a specific methodology in which Joan is especially well-versed, or even better that she has contributed to create or advance; it is usually not difficult to make a case for productivity. Most important is to convey the notion that Joan and I have worked together, that I have interacted with her regularly, observed her “in action” — especially if I have a large group, one may wonder how well I really get to know my own advisees.
If I feel that, on any of the above three points, Joan may not appear on paper as strong as I wish, this is the place for me to explain (again, in as quantitative as possible terms) why the record fails to reflect Joan’s real value .
I am then commenting in more general terms on aspects that I regard as (usually) less important, such as the candidate’s ability to deliver a good, clear presentation, her written and spoken language skills, her ability to work with others and so on .
Because I want you to take my letter seriously, I am not giving in to the temptation of using hyperbolae, of comparing this candidate to some famous physicist or describing her in unrealistic and borderline grotesque terms (“she’s the next [insert name of favourite legend here]”, “the brightest young scientist whom I have ever met”, “simply the best around”, “sometimes I suspect she is superhuman”, “walks on water” etc.). I am pleased to have observed in recent years that that habit, which became commonplace during the disgraceful 90s, has largely faded, and for the most part letters have gone back to being refreshingly sober.
They would be good for you
Not only is Joan good, she is a perfect candidate for you. Her interests in ThatSpecialProblem or ThatOtherImportantField make her a very good complement for your department, which is aiming at building a strong group in ThatGeneralArea, witness your recent hires of John Super and Mary Excellent, with whom Joan would surely profitably interact. Furthermore, she would bring to your department her world class expertise in ThatGreatTechnique, which is proving quite versatile and powerful.
I am obviously in no position to assess Joan’s teaching ability, as her teaching experience is limited, much like for any scientist at her stage, who by and large rightfully focus on building a strong research portfolio. However, I have no doubt that Joan will make an outstanding instructor. She delivers very clear, well-organized and entertaining presentations, has a warm and friendly disposition, and is genuinely interested in engaging others in her work. She has been serving as a very effective mentor to my graduate student Eugene Drack, for example, as well as supervised two undergraduate students last Summer, one of whom is now wanting to pursue a doctoral degree working in my group. Last but not least, Joan has taken the time from her very busy work schedule to participate to two major outreach initiatives during the past few months. I was not directly involved in them, but it is my understanding that her contribution has been highly valued.
OK, now, size ’em up !
OK, now we have come to the point that I hate the most — but I understand that it has to be done. After all, I do know who Joan’s main competitors are — it is almost inevitable, as there are not actually that many around in that area, looking for a position, and it is not implausible for someone sufficiently senior to know (of) most/all of them. And let us not kid ourselves, I am not writing a letter for Joan only. Frank Glotz and Bill Grump, also postdocs in my group, as well as Louise Blink, my former graduate student, who is now a postdoc at ThatOtherPrettyGoodPlace, are also applying for this very same position. They may be all good and capable of the job, but I know that you expect me to give a frank opinion as to who among them is the best — the person to whom an offer should be extended first.
Given that that is the expectation, like it or not, it would be ridiculous and counter-productive of me to say “they are just all equally exceptional and I shall not compare them”, as some do. By doing that, I am most likely to undermine my objectivity and the credibility of all of my letters, and as a result weaken the case of all of the people for whom I am serving as a reference.
So, given that if I have accepted to write a letter, and I am speaking positively of all of them, I regard all of them as good choices, I believe that Joan is a notch above both Bill Grump and Louise Blink, as well as technically stronger than Laura Klupp, a recent assistant professor at BetterThanYou University and more original than John Zatt, at SameLeagueAsYou. Of all the young scientists whom I have had the pleasure of working with in the past decade, Joan is probably the third best, second only to Anna Amazing, currently an assistant professor at BigTime University, and Frank Glotz, whom I see as slightly more driven and efficient than Joan. For this reason, if I were to make a choice between Joan and Frank, I would go with Frank. I understand that there is a fair degree of subjectiveness in this judgment, which is why I would recommend (if possible) that you invite both of them for an on-campus interview, and make your own independent evaluation, or at least arrange for an informal conversation at the upcoming Huge Meeting, which they will both be attending. Let me restate, however, that I do believe Joan to be an excellent candidate for this position.
Please feel free to contact me directly, should you require any additional information or wish to discuss the above matters with me in person.
 It is quite common to read letters written on behalf of candidates whom the reference does not really “know”. This can happen for a number of legitimate reasons (one ought not forget that by writing letters, references put their reputations on the line). The most common is that a junior scientist simply may not have as many close references as letters required for a particular position, and might resort asking a person who mostly knows of her. There is nothing wrong with that per se, but readers will be able to tell, and the letter will typically carry less weight. Sometimes, a scientist who is not the candidate’s direct supervisor (at least on paper), may know the person as well as (or even better than) the supervisor, by virtue of having collaborated closely with the candidate. If that is the case, then the reference has to explain in detail the circumstances of their interaction, which may make this letter (at least) as important as that written by the direct supervisor.
 The reference of course must assume that the letter is read in the first place, i.e., that the committee is willing to give Joan a chance. After all, her degree is not from one of the few well-known universities that are known to be vastly better than all others (for the simple reason that everyone says so and acts consequently, and therefore it must be true. We all know that, and so I shall not waste any time repeating truisms).
 What is the point of stating the above ? In my view, it is important that a future assistant professor be able to work independently. She will be expected to start a fresh, innovative research program, capable of generating interesting projects on which to engage graduate students, of sustaining itself by attracting the needed extramural funding, and generally of raising the profile of the department. It has been my consistent observation that that correlates reasonably well with a person’s overall curiosity and desire to explore different avenues of inquiry.
I have seen countless examples of graduate students who sported an impressive publication record as freshly minted PhDs, but failed to continue at the same level in their postdoctoral appointment, not enjoying the same degree of supervision and perhaps working in less nurturing an environment.
I have also seen cases of scientists who made significant contributions in one area, typically during their graduate studies, and remained almost wedded to that area (or even specific problem) for the rest of their careers. What is the (potential) problem with that ? Well, sooner or later even the richest research field dries out (or falls out of fashion), at which point the fortunes of an investigator unable to change focus make a turn for the worse — funding is typically the first thing to be lost, followed by graduate students.
So, to me it is an important observation that a junior scientist has been able to switch area of research after completing his/her doctoral studies.
 Unless of course that happens to be the truth, in which case that is a diplomatic, and actually quite effective way of conveying the message.
 This is less of a problem for experimentalists — one cannot really spend time in a lab without doing or learning how to do anything useful. When it comes to theorists, however, often times one runs into some who have been trained as “philosophers”, rather than scientists. They are very good at “thinking outside the box”, “conceiving novel scenarios”, “exploring the deeper connections”, often without doing a single (non-trivial) calculation. Their
wild bold speculations hypotheses are based on stunningly simpl istice models. Thy are not interested in making actual, testable predictions, no… they leave that mundane task to less visionary colleagues, perhaps those who can actually do stuff more “technically inclined”, who are lucky enough to try to make something out of build on the bullshit deceivingly flimsy, elusively subtle foundation laid out by their superiorly talented peers.
I have no advice to dispense to anyone wanting to write a letter of recommendation on behalf on these gifted young scientists — first off, they typically do not need it, secondly they are simply much too deep
into it for me.
 There are two schools of thought, here, one according to which it is best not to mention a perceived weakness, as one may simply give a reader who would not have otherwise noticed it a reason to be concerned. The other school, toward which I lean, holds it that it is best to lay all the cards down and be preemptive, acknowledge an apparent weakness and explain why it ought not be cause for concern, and why we are still being so forceful in promoting the candidate. I suppose, however, that it depends on how large the hole to fill is — if it is not, one might as well not even mention it. Of course, that is a judgment call, one of the many in this highly imperfect endeavour.
 Why are those things “less important” ? Well, clearly a person with unusually weak communication or interpersonal skills (in the perception of the writer, of course) is not going to make a good choice for the position. In fact, given their obvious implications on that person’s likely effectiveness as a teacher, if her deficiencies in this area are particularly significant, that may well trump anything good one may have to say about her, from the scientific standpoint. However, my impression is that such cases are rare; in the vast majority of cases, someone who has made it through graduate school, gone through one or two postdoctoral appointments, built a respectable research CV and given invited talks at conferences or departments across the continent, knows how to deliver a presentation and speak intelligibly. Of course, if that is not the case one ought to warn the reader, but at that point it may be better for the reference to talk openly and bluntly to the candidate about this type of problem, even before the time comes to write a letter.