Mathematics & Computer Science Staffing Overview

Tim McLarnan
Mic Jackson
Joy Williams Lind
September 18, 1999

 Staffing Levels

Mathematics

    Because of the amount of shift and confusion in staffing on the second floor of Dennis in recent years, it may be useful to start with a large view. Ten years ago, the Mathematics Department was staffed at a level of roughly 24 courses or 80 semester hours per year. (This level was in fact stable for a number of years, and is not an attempt to pick the richest year in the Department's history. In fact, if we were willing to go back slightly farther, we could get the level up to at least 28 courses or 93 semester hours.)
    Since then, a variety of events including Dick Rodgers' growing commitment to Fine Arts, added commitment of everyone in the Department to Computer Science, and a collection of fractional hires and fractional retirements have resulted in a current staffing level of 63 semester hours in 1999-2000, a roughly typical recent year.
    In the absence of additional hiring, the Math Department will be at about 53 semester hours in 2000-2001.

Computer Science

    Over the same 10 year period, Computer Science has remained quite consistent at very nearly 11 courses or 33 semester hours per year. In the current 1999-1900 academic year, we are offering 10 courses on budget, and Charlie Peck is teaching 4 unpaid internships.
    In the absence of additional hiring, CS will be staffed next year at roughly the same level as this year: by 3 courses from Charlie, 2 from Tim, 1 course each from Welling, John Howell, Lew, Ray Ontko, Ed Delaney, an overload course shared by Joy and Ray Ontko, a May Term course shared by Charlie and Ray Hively, and 3 free internships (each with multiple students) supervised by Charlie.

Current Courses and Wishes

Mathematics

    As we think about the appropriate staffing level for Mathematics, it may be helpful to try to break the Department's curriculum down into two categories of classes. First, there is the A List: our irreducible core. Eliminating any of these would mean either that we could not offer critical service courses or that we could not continue the mathematics major:
 
 Sections  Total hours  Class
3 
9 
  Elementary Statistics *
3 
15 
  Calculus A *
1 
5 
  Calculus B *
1 
3 
  Discrete Mathematics *
1 
4 
  Multivariate Calculus
1 
3 
  Differential Equations
1 
3 
  Linear Algebra
1 
3 
  Algebra A
1 
3 
  Analysis A
1 
3 
  Algebra/Analysis B (alternate years)
1 
2 
  Seminar
 15 total  53 total

    The next set of classes, the B List, represents what we think are either highly desirable service courses or courses which address some of the glaring weaknesses of the major above or courses which build on important strengths of the Department and its faculty to keep mathematics at Earlham a program which serves and attracts students. We would teach all these courses if we could; currently we either do not teach them at all or teach them on at most an alternate year basis. A fair view of a typical year in the Mathematics Department is that it consists of all the courses in the table above, plus 40% of the courses in the table below:
 

 Sections  Total hours  Class
1 
3 
  Mathematical Modeling
1 
3 
  Geometry (*?)
1 
3 
  Intermediate Statistics
2 
6 
  Topics
1 
3 
  Math/CS Topic
2 
6 
  Operations Research
1 
3 
  Symbolic Logic *
 9 total  27 total

    (For a description of why we think these courses are important, please  click here .)
    It is probably not entirely coincidental that the sum of the courses in these two tables amounts almost exactly to the historical staffing level of the Department 10 years ago.
    Of course, there are other courses we would like to add beyond these. ("If wishes were courses, professors would ride.") For some of us, College Algebra tops that list; others of us believe that we cannot offer this course for the slightly paradoxical reason that we think the demand is so high that even trying to meet it would require additional staffing.

Computer Science

    The Computer Science program consists of the following courses offered annually:
 
CS Applications for a Global Society *
Programming and Problem Solving *
Advanced Programming
Algorithms and Data Structures
Principles of Computer Organization
Theory of Computation
Senior Seminar

and of the following courses offered on alternate years:
 

Principles of Programming Languages
Operating Systems
Topics - Database Systems
Topics - Software Engineering
Topics - Cognitive Computing
Topics - Robotics
Topics - Parallel Computation

Charlie also leads a number of internships in Applied CS, doing things like writing WebDB. There have also been some courses like Parallel Computation offered by Joy and by Ray Ontko as free overloads.
    If we could add courses in CS, we would convert some of the alternate year courses like Languages and Software Engineering to annual offerings. We would also add courses in Object-Oriented Software Construction and in Compiler Construction.

Hiring Option 1: 1 FTE in Math

    Superficially, the situation here would seem to be simple. Computer Science is adequately staffed; Mathematics plainly is not; so we should make a full-time hire in Math. This would bring Math up to about 70 semester hours next year - up 7 from this year, down 10 from 10 years ago. This would mean that Mathematics could offer more but still not all of the courses in our wish list above.
    What a hire like this would mean outside the Mathematics Department and the CS Program is unclear in detail. Some of the courses we would wish to offer (like Modeling, Operations Research and Intermediate Statistics) are clearly of interest outside the Department; others are not. We have tended to make decisions about which of the courses on our "B" list get taught as we prepare unit plans, but we recognize that the College could elect to mandate certain choices from this list.
    In particular, CPC asked last week specifically about Symbolic Logic, and about the ability of Mathematics to contribute to Environmental Science. Our view as a Department is that even with a 1 FTE hire in Mathematics, we could not undertake either of these tasks without pain. Both would involve trade-offs of not teaching classes we think are important to other Departments or to our majors. Doing either would hurt the Math Department. Our current opinion is that the benefits to the institution as a whole of Mic taking leadership of Environmental Science may outweigh the loss to Mathematics of 1 course from him, and that the advantages to the College of Symbolic Logic as an annual course do not outweigh the costs to the Department of offering it; but those decisions have not yet been made. (We thought agreement about Logic had been reached with the Philosophy Department several years ago, but this seems not to be the case; a meeting is scheduled Tuesday morning that may or may not resolve the question.) The fact remains, however, that a 1 FTE hire in Mathematics is not a magic solution in which the needs of Philosophy and of Environmental Science can be met out of a vast surplus in Mathematics.
    There are additional problems with doing a full hire in Mathematics now. One is that the unit plans outlined above assume Tim McLarnan teaches half time in Mathematics and half time in Computer Science. This has not been true until this year, though, and it is not going to be true indefinitely. Tim was hired in the Math Department with no commitment to teach CS. Over the years, he has agreed informally to teach increasing amounts of Computer Science, but he has not made long term commitments to do so indefinitely, and he is not making those commitments now. He intends to return to teaching more or less full time in Mathematics. The result is that hiring now in Mathematics entails a commitment to add additional staffing in CS as Tim returns to Math; and it may not be easy to find half an FTE worth of adjuncts or faculty from other Departments interested in teaching upper-level CS courses. Assuming that this staffing will be easy to secure may well be a mistake.

Hiring Option 2: An Amphibian.

    So now the optimal solution would again seem to be obvious: either hire a single person who will be 50% Math and 50% CS, or make 2 half-FTE hires in the 2 Departments. If Tim then returns to Math, the staffing of both programs is exactly what it would be under the previous plan, and everybody is happy.
    The fundamental problem with this plan is that we probably can't do it. People comfortable spending their careers doing upper-level CS classes and teaching 3 math classes a year are probably much less common than folks preferring to be anchored in one Department or the other. Our odds of hiring the perfect amphibian seem slim. Equally well, the chances of finding 2 people both interested in half-time positions at Earlham in Math and in CS do not seem overwhelmingly good.

Hiring Option 3: A Mathematician Who Can Teach Some CS.

    So why not be more moderate and go for a Math hire who can teach 1 or 2 courses in CS? Probably we can persuade Tim to keep teaching a course or 2 in CS, and lots of mathematicians know at least some CS and could teach a course or 2.
    There are two problems with this scheme. One is that the courses Tim has been teaching are no longer courses close to mathematics that any mathematician could step into with ease. They are hard and central upper-level CS classes. There's no guarantee, in other words, that this plan would really be much easier than hiring the amphibian.
    The more serious problem with this plan, though, gets us finally to the curricular core of the proposal that we hire a Computer Scientist. Computer Science at Earlham is a program which is spread among too many faculty and which is too dependent on one person. CS consists of 1.8 FTEs spread among 8 or 9 or 10 people. This means that there is only 1 person at Earlham whose central focus is Computer Science - Charlie Peck. Charlie teaches 3 courses, convenes the program, advises the vast bulk of the students, and spends enormous amounts of extra-curricular time on Applied CS projects. A growing and vibrant Program that graduates more majors than Math or Physics with smaller staff than Geology and with only one central figure - himself half time - has to be seen as dangerously narrow. To be as blunt as possible, if Charlie were to elect to leave the College or if, God forbid, he should die (killed by an irate customer at Infocom, perhaps, or gored by one of his sheep) then Computer Science here would be in grave trouble. We would need instantly to make a hire, since much of what Charlie does could not be done by anyone else at Earlham. There is no way we could make a half-FTE hire in CS and replace Charlie, and there is no guarantee we would be successful in a 1 FTE hire. Without that hire, though, we would have grave difficulty sustaining the major; and one shudders to think of the implications of telling 15 or 20 juniors and seniors to find new majors. CS at Earlham is a great program getting better and better and attracting more and more students; but it is a program resting on the back of one irreplaceable person.
    Hiring a mathematician who can teach a course or 2 in CS would turn a program consisting of half of Charlie, half of Tim, and some odd bits, into a program consisting of half of Charlie and some odd bits, with no projected retirements to alter that situation until perhaps 2015. This cannot be wise.

Hiring Option 4: A Computer Scientist.

    This is the process by which the Mathematics Department moved to its current position of advocating the hire of a computer scientist. A second computer scientist could, in our judgment, do an enormous amount to strengthen and stabilize Computer Science at Earlham. Right now, Charlie's legs are holding up the whole program, and as we say in the trade, "Four legs good, two legs bad." (Six legs would be better still, but we're realists here.) For us, this is the overwhelming curricular virtue of hiring in CS: we are safeguarding the future of a vibrant program, broadening its intellectual base, and enhancing its external credibility. These are terribly important institutional goals.
    At least two questions remain, though: can we realistically expect to make a hire satisfying these goals, and what are the implications of such a hire for the needs of the Math Department?
    With regard to the question of the possibility of a CS hire, we can't claim that hiring a computer scientist will be trivial. We routinely graduate students in CS who make starting salaries higher than the current salaries of anyone in the Math Department, and a Ph.D. in CS can obviously command much more than that. Informal mention of the possibility of a position here has attracted at least one interested party, though; and other GLCA schools have hired CS faculty. Positioning Earlham as a school at which a computer scientist can be a community member rather than an employee may be attractive to some people. The chance to be a formative agent in a small but dynamic department experiencing rapid student growth and possessing a very diverse student body may also appeal to some. Moreover, if we start now, and not when something awful happens to Charlie, we can afford to be patient. There are lots of mathematicians on the market, and if need be, we can certainly go through a cycle or two of failing to hire in CS and then making temporary hires in Math and trying again. Current faculty are prepared to staff CS at current levels while we carry on this search. Although one or more temporary hires in Math might not be great for the Department or the College, we think we should be able to find folks on that basis who can keep the Department running and staffed at the same level as Option 1 (the 1 FTE Math hire) above.
    If hiring a computer scientist is to work, though, we have to be able to keep both CS and Math staffed at essentially the same levels they would be at if we made a 1 FTE appointment in Math and if Tim agreed to maintain a 50/50 split between the two Departments in perpetuity. How do we turn a CS hire into Math staffing? We have several real and hypothetical approaches to this problem:
  1. A full-time CS hire would free Tim to return full-time to Mathematics immediately, a net gain of 3 courses for Math. At this point, CS is up 3 courses and Math down 3 relative to the plan of a 1 FTE Math hire.
  2. We would like to advertise for a CS position with the addendum that a willingness to teach 1-2 courses in Math is a plus. We may not find anyone with this willingness; but then again, the one person we know is looking at the (as yet hypothetical) position not only could teach in Math, but is writing a book on Linear Algebra for Computer Scientists; so such people clearly exist. Finding such a person would move an additional 1-2 courses from CS to Math, bringing us in the best case only 1 course away from the staffing mix resulting from a 1 FTE hire in Math.
  3. Physics faculty currently teach 2 classes a year in Computer Science. The Physics Department is now discussing whether they are interested in the possibility of instead offering 1 or 2 courses a year in Mathematics (in addition to Applied Math/Theoretical Physics, which the Department already contributes). That discussion is ongoing, but the initial reaction of both Math and Physics has been some interest in the idea of sharing courses between the two Departments. There are ways in which this may actually be an easier and more natural collaboration than that between Physics and CS. Of course, the choice to make this trade is entirely up to the faculty involved, and they may in the end elect not to be willing to trade.  If they are willing, though, the possibility exists to move 1-2 more courses from CS to Math.
  4. One other possibility for moving staffing into Mathematics involves adjunct faculty. This year, Bob Hunter has agreed to teach one section of Elementary Statistics as an adjunct. Given Bob's experience using statistics as a social scientist and his deep expertise with computers, we as a Department are enthusiastic about this experiment. Should Bob enjoy the experience and be an effective teacher, then we might explore continuing this relationship in future years. This could be done by replacing a CS adjunct with Bob, or by adding Bob in exchange for 1 course from Mic in Environmental Science. (The current adjuncts in CS are Ray Ontko and Ed Delaney. Ray seems interested in a long-term relationship with the College, and has been active teaching courses on overload (as an adjunct!), planning curriculum, and donating and installing hardware for the Department. It would be a mistake not to continue to take advantage of his energy, expertise, and enthusiasm. It is less clear that Ed intends to teach with us long term, which might make a trade for Bob relatively socially painless as well as curricularly desirable.) One could even imagine Bob teaching one section of Stats each semester, instead of one each year; and one could imagine hiring an adjunct other than Bob for this purpose.

    5.  

     

        Apart from the first item above, none of these possibilities are sure things; but together they amount to much more than the transfer of 6 courses from CS to Math that would be necessary to give staffing equivalent to a full Math hire. To us, it therefore seems prudent to start a search in CS and to work to make the swaps above a reality. It might be prudent to reach an explicit initial understanding about what the minimum course transfer from CS to Math is in order to justify a CS hire. The Math Department would be comfortable losing one course in the process, but would probably oppose a hire in which the net gain to Math was only 4 courses instead of 6.

    What if CS Gains 1 Course and Math Loses 1?

        I hope the foregoing material addresses some of the concerns of CPC, but it is possible it is only introductory, in which case I apologize profusely. A major concern of CPC on Tuesday appeared to be what would happen curricularly if we hired in CS and could not move 6 courses from CS to Math. We hope never to find out, and we think we have enough ways to move courses that we will not have to find out. If we fail, though, then Math will gain one less course from our B list than we would otherwise. We can't yet say which course we would give up, and it might not be the same course each year. The number and interests of majors fluctuate each year, and we respond to those fluctuations. Right now, we think it unlikely that we will be teaching Logic in any case; which would mean that whatever course we drop will not be a Gen Ed course, and that it will not be picked up by any other Department.
        The tighter staffing is in Math, as well, the harder it will be for us to justify losing additional courses to Environmental Science. Perhaps, though, there are additional ways to add courses to math than just by taking them from CS. One could hire adjuncts, for instance.
        If Math loses a course, then CS gains one. What are the curricular implications of this addition, or how should it be spent? The Math Department doesn't have an opinion on this issue. It would not seem unreasonable simply to allow CS, which graduates far more majors per faculty FTE than the College average, to add a course. Probably this would mean converting an alternate year course to an annual course, simplifying student scheduling and making it easier for our majors to participate in off-campus programs. The case could also be made that the College would be better served by leaving CS at 1.8 FTEs and letting John or Lew teach one more course in Physics, or Welling teach one more course in Politics, or by replacing an adjunct in CS by an adjunct in some other discipline. This is a choice among good things, though - any of these choices is a benefit, not a cost.

    Finale

        In summary, then, attempting to hire in CS now gives us a chance significantly to strengthen and to stabilize a successful and growing program. It maximizes our chance to make an eventual hire in CS by starting early, and by trying to make a full-time hire. These seem very substantial benefits. If we are able to shift resources from CS to Math - and we think we will be able to do this - then choosing to hire in CS instead of in Math has essentially no effect on Departments outside the 2nd floor of Dennis. Indeed, it frees resources for them, since it doesn't commit us to add in CS as Tim moves back to Math. If we end up with a shortfall in Math and a surplus in CS, then the costs outside Dennis again seem rather small to us unless this means the difference between Math's participation or non-participation in Environmental Science. With good will from all concerned, this can probably be avoided.
        We cannot and should not mandate that a CS hire take place regardless of the effects on other Departments; but to permit the residents of the 2nd floor of Dennis to attempt collectively to meet the staffing needs of their programs by starting with a hire in CS and continuing to work together still seems to us to represent good policy.