This morning I am going back and starting over, with a part of the collection. Since mid-July I've been entering information from Dennis Brett's card index into the database. Once entered, I then planned to begin going through the 100+ boxes of specimens, filing them away in our newly-obtained cabinet system, and noting that new location in the database.
But last week, while grinding my way through the cards, something clicked and I suddenly realized I was messing up my count.
Let's back up a few years. When Mr. Brett donated the collection, he asked for an appraisal for accounting purposes. After checking the options, Penn State decided that I should be the person to perform the appraisal. And I spent the better part of six months doing so.
But in the process, I came across a slight discrepancy. When I first started pondering the size of the collection, I measured the stacks of cards in Mr. Brett's files and estimated that there were 5,000 to 5,500 in the collection. Dennis had performed the same estimation process and concluded that there were 6,000 specimens in the collection. Mr. Brett had been meticulously filling out an index card on every specimen in his collection as they were collected over the years.
|Dennis Brett typed out a lot of cards over the past seventy years. Those four cabinets are jammed full.|
For the actual appraisal, I didn't use the cards at all, since I needed to examine the actual specimens. Instead, I went through the entire collection, box by box, selected a random specimen from each box, ascertained a value for the selected specimen, applied that value to the number of specimens in that box, and then added the estimated values of all the boxes to arrive at an estimated value for the collection. Over the course of this process, then, I actually determined the true number of specimens in the collection...which was "only" 3,983.
So both Mr. Brett and I had apparently overestimated the number of specimens in the collection by "measuring" the cards. I had no apparent answer for the discrepancy, except that the "measurement" was rough and perhaps a small error could have been multiplied. Dennis also doubted the discrepancy, and felt certain that the actual number was closer to 6,000 than 4,000. After all, he had spent 70 years building the collection, so I took his viewpoint seriously. But...I had counted every piece...could I trust my own tally? I decided to wait until I had the whole collection documented, and the final number would be confirmed. Then I forgot about the issue.
That was back in early 2016. So, by mid-2017, when I started entering the information from the cards into the database in the final documentation effort, the specimen tally issue was really not in the forefront of my thinking. After all, I was ultimately going to match every specimen to a data entry, and reconcile any missing specimens, or eliminate any data entries for which there was no specimen, at the end of the process.
So, I started on the "A's" beginning with Abies amabilis, entering all the data on each card in the appropriate cell in the database. When I reached Abies lasiocarpa, I noticed Mr. Brett had entered a synonym name, A. subalpina. I confirmed that Abies lasiocarpa is the accepted name, entered that in the appropriate column, and moved to the next card...which was for a specimen also labeled A. lasiocarpa, with a synonym of A. subalpina.
This is where fate led me down the wrong road. I briefly pondered if these were the same specimen...but then noticed that each was from a different location. The first was from Oregon, the second from Colorado. So they were indeed different specimens. I made the mental note that Mr. Brett had occasionally entered synonym data on the cards, but made the erroneous assumption, based on the confirmation of different locations in this first instance, that each card represented a unique specimen in the collection.
The fate that worked against me was that these two cards preceded the next two...which were for, I assumed, two specimens of corkbark fir. The first was filed as Abies lasiocarpa var. arizonica, with a synonym of Abies arizonica. The second was for an Abies arizonica, with a synonym of Abies lasiocarpa var. arizonica. I confirmed that the accepted name for the species is A. lasiocarpa var. arizonica, and entered both into the database. As I did, I apparently didn't give enough thought to the source of both specimens...Arizona.
After all, you could assume that a species named A. lasiocarpa var. arizonica would mostly be found in Arizona, so not unusual that two different specimens might be from that state, right? I forged ahead.
But two months and nearly a thousand entries later, it finally dawned on me. By now you've probably figured out my problem, especially if you've built your own card catalogue. Mr. Brett had created multiple cards for those specimens which were labeled with synonym names, so that he could check them against future possible acquisitions under both (or multiple) names. Makes sense.
But it also makes the number of cards in the index larger than the number of actual specimens. In this case, about fifty percent larger.
So, if I was just entering a few hundred cards, I would just keep plunging forward and weed out the multiple entries as I check them against actual samples. But a few thousand? When it takes me about a month to enter roughly 500? Forging ahead at this point would be roughly akin to jumping off the cliff, just because I happened to walk up to it.
Besides, it causes a real problem with my real-time reporting of my progress. If you look at the box on the right, you'll see that I'm reporting that there are currently 5,176 documented specimens in the collection. That number is high...but how much, I don't really know. Meaning I have to go back and clean up the database by deleting the duplicate entries as I pull them from the card file.
So, this morning, I'm doubling back and cleaning up the mess. First, I'll go through the cards I've entered, and as I pull their duplicates, I'll delete those entries from the database, and correct the tally as I go. Then I'll continue through the entire catalogue, until I pull all the duplicate cards. I'll set these aside for later.
Finally, then, when completed with this data entry (which will be two or three months earlier than it would be if I entered all the duplicates) I'll go through the boxes, pull the specimens and match them to their entries. If any of the pulled cards are actually different specimens, I'll find those cards in the stack that I've pulled, and enter the data.
Yes, it sends me back to "Start" today...but it means I'll "Get out of Jail" earlier.