My memories of Karl
Contribution by Phil Williams
My memories of Karl stem partly from the many conferences that we have both attended, and partly from the memorable year November 1976-Novenber 1977, when I was privileged to work with him in his laboratory in Beltsville, MD. The conferences included the IDRC conferences, held every even year at Wilson College, Chambersburg PA, and the international conferences, which rotated among countries each odd year.
Phil (as he is!)
During the second IDRC conference in 1984, Karl and I sat with Gerry Birth, the founder of the IDRC conferences, and discussed the concept of a monograph that would assemble every aspect of NIRS. The monograph was duly assembled, with contributions by experts in all of the aspects. At all of those conferences, beyond greetings, I purposely stayed away from him. He was continuously surrounded by NIRS workers, young and seasoned, who sought his knowledge, and advice, and who only had the opportunity to share his wisdom on those annual occasions.
So my memories of him at conferences were mostly of him in a head-to-head huddle with someone, or with a small cluster of acolytes. An excellent example of his benevolence occurred during the memorable 1986 conference in Hungary, when the ICNIRS was founded. One afternoon there was a sight-seeing bus-trip. On the way out I was sitting beside a young lady, who asked me questions about application of NIRS to determine apple quality. After we had discussed the things that she could do (including identifying the factors that affect the variance in apples, and assembling sample sets), she mentioned that she would like to have met Mr. Norris. At the first stop I introduced her to Karl, and told him that she would like to ask him some questions, She was with him for the rest of the trip, and I’m sure that she has the same sort of memories of him that I and many others have experienced.
He and I carried out our conversations by telephone, whenever the occasion demanded, usually whenever I needed to know something about the spectroscopy. Our conversations were short, because he always had the answers, and both of us were busy. I had the privilege of working with him for a full year in his lab in Beltsville MD. During the first 1 or 2 days we talked about the evaluation of calibrations. He showed me his results on prediction of the protein content of wheat, using his second derivative quotient math algorithm. He had combined 6 sets of wheat samples, and had obtained a SEP of 0.066, after 8 iterations. I asked him what would happen if he calibrated on 5 sets and predicted he other one. He said that the result would be the same, so I asked him if we could try it anyway and we obtained a SEP of 0.27. He was a bit surprised at that, but I said that that was probably the right result. Then we returned the predicted sample set to the combined calibration set and repeated the procedure, using another set, and this time the SEP was 0.26. I pointed out that to evaluate a calibration it was necessary to predict a set that had not been used in the calibration of which, to my surprise he had not been aware.
He arranged for the head statistician of the USDA to come to talk with us, and he, of course said that that was indeed the true method for evaluation of a calibration. In other words, Karl had been using the SEC all along, up to then. So Karl actually learned something from me, but I would like you all to know that he learned something from all of you who were privileged to talk with him. You all contributed to his wealth of knowledge.
During our year together we addressed the variables of particle size and moisture content in predicting protein content of ground wheat. I designed the experimental work, and arranged for all of the samples to be assembled, and sent from the GRL, Winnipeg, including all of the laboratory analysis. My colleague, Dr. Susan Stevenson, did a great job on assembling the sample sets for us. I set forth to do the calibration work, using Karl’s Cary 14 spectrometer and his quotient math software. The first thing that I learned from Karl was how to access his computer and its software. I got my ball-point pen and a note-pad, and took notes from Karl. When I was through, I had 12 pages of notes on which switch to toggle and in what sequence, and what the screen should say. It took about 3 minutes to get his system running. From time to time in the early days I would make a suggestion about a small change that would make things a bit easier. Next morning, I would be chagrined, to find that I couldn’t get the computer up. He would emerge with a sly smile, explaining that he had made the change that evening and changed the toggling process!
Eight months later in August, when the calibrations were all complete, we sat down to see what we had achieved with his quotient math algorithm. After testing 6 or 7 calibrations, and used them successfully to predict protein content in samples that differed very widely in particle size or moisture content, using reciprocal calibrations, Karl quietly said “I think we have a breakthrough”. That moment will stay with me forever.
Later that year (1977) we introduced these applications of his quotient math pre-treatment at the AACC annual conference in San Francisco. I was spellbound by the diversity of his vast knowledge of spectroscopy, from the ultraviolet to the near-infrared, as well as his knowledge of the voluminous aspects of operation of the equipment, all of which he enjoyed to share freely with anyone who sought his expertise. But there is more in life than NIRS, and during the year I learned about Karl Norris, the Man, his subtle sense of humour, that would emerge unsuspectingly from time-to-time. Also his fondness for his monthly poker evenings with some of his colleagues, to which I was invited. His remarkable memory often came to his aid during those evenings, to the frustration of the rest of us! Truly a memorable year, and the founding of a collaboration and friendship that persisted for 44 years.
After his retirement I called him every few weeks, and our conversations were a lot longer than during his working days. He grew very sad when he began to lose his memory, and could no longer use his software. We continued to enjoy our telephone conversations until only a few days before he left us, on July 17th, 2019, and bless him,
He remembered my voice right up to the end!
Phil Williams
August 6th, 2019.