Chapter 3: Epigraphy and Teaching

My morning did not go well at all. I was planning to spend time before classes on examining the journal. I had barely started my first cup of tea when Caroline called.

"Mitch? I have some good news. I was offered a job at the insurance company that I was hoping for. I will be managing their short-term capital!"

Caroline was teaching at the Business School. I knew she was hoping to find a job in the industry, but I knew nothing about insurance companies.

"Congratulations! Where is the job? And when do you start?"

"Well, that's the bad news. The job is in New York City. I start in three weeks, but I am taking the rest of the week off to go there and look for an apartment."

"What? You'll be leaving so soon?"

"I am sorry, Mitch. It's a dream job for me. It sucks that there is so little time, and I wish I could discuss it with you in person, but I am flying out tomorrow and there are so many things to take care of. I am sure you understand..."

I didn't understand. Caroline and I had been going together for a year. I thought we had a pretty good relationship. And she was telling me we may not even see each other again?!

Talking didn't really help. Caroline was excited and in a hurry, and I was sounding whiny even to myself. I extended all the best wishes and hung up still on good terms. What was the point in throwing a hissy fit? I did wish to throw the mug against the wall, but I thought my neighbors wouldn't appreciate it. Not to mention that the poor mug was not at fault.

I poured the lukewarm tea into the sink and wandered around the campus until it was time for class.

"Professor Acker?"

A pretty, older redhead stood in my office. A graduate student?

"Yes. How may I help you?"

"My name is Jenny Sanders. I would like to take Language Change 302, but I haven't taken 301. I studied a similar course at Lincoln State several years ago and I think we covered most of the syllabus. The Registrar said I could take it with your permission."

"What degree did you pursue at Lincoln?"

"History. I got a BA there, but now I am working on my Master's in Anthro here at Coleridge."

"Okay, you should be able to test out. I can prepare a test for you by Friday; will you be ready then?"

I saw her wince. She was probably hoping to avoid testing.

"Don't worry, as long as you get at least 50 percent I will let you in. I will not expect you to be perfect, and the grade will not be recorded. I will be here on Friday from 1 to 3; you can come and take the test – it will require 45 minutes."

She looked a bit more cheerful, and after writing down her name and contact info I wrote her in for 2pm on Friday.

I work at Coleridge University. I teach in the Department of Anthropology. The study of mankind is a huge field, with many branches and interactions with other disciplines. But I am an epigrapher by specialization and passion.

Epigraphy is the art of decoding inscriptions. Epigraphy doesn't concern itself with the meaning of those inscriptions; that's a job for historians or scholars of literature. Still, with all human languages you can't separate syntax from semantics. The disciplines and specializations are artificial boundaries; to translate a culture's writing one must understand something about the culture.

I call it an art, because while many scientific approaches are used in analyzing ancient writing, at the end of the day it's human intuition, or brilliance, or sheer stubbornness that gets them decoded.

On the other hand, I consider myself a scientist. Thus, I am a scientist practicing art. Of course, art itself is a word that changed meaning fairly recently, which can easily divert me to talk about languages – another important aspect of anthropology as a whole, and epigraphy specifically.

All of which, of course, made the journal all the more fascinating. Not only did it have that tinted writing, but it had some hitherto unknown writing on it. Once again, I considered that the alphabet was far more recent than the presumed two millennia that Colt mentioned.

If so, it was a pretty elaborate hoax. In any case, I wanted to understand those twenty five pages, and figure out the graphs – if they actually meant something.

I read the journal again, this time more carefully. Colt was not able to make anything of the strange characters, at least as far as this journal documented. Once again I resolved to try finding his later records.

In the first half of the journal, Colt had several drawings of sites and artifacts. He had a very fine hand, and he would be welcome on any modern dig.

The strange writing remained enigmatic as ever, but I started to look at the characters themselves and their position. The first page or so was a bit shaky, but it looked like Colt got used to their shapes and they became more regular. The letters were intricate, and vaguely rectangular in shape. After the first page they seemed to be rigidly aligned, like typewritten text, but with no obvious spaces or punctuation. Once I realized that, the first page showed the same grid, but the columns were shaky, possibly due to Colt's lack of experience.

I examined one page closely. While the characters – I still did not know if they were letters, syllables or entire words – were similar, no two were identical. This was going to make decoding more difficult, but the puzzle was only getting more fascinating. Deciphering ancient scripts is very different from decoding encrypted text. The ancient letters can be easily misinterpreted due to damage, as well as from not recognizing what variations meant. Sometimes identical letters on ancient tablets looked very different, and sometimes different ones looked the same. What was the significance of a dot in the upper left corner, or a curved rather than a straight line?

On the other hand, deliberately encrypted text usually had clear characters, but no recognizable patterns. The decoding techniques were vastly different.

And yet, under a series of mapping rules, a bad interpretation of ancient texts, and failing decoding of cyphertext, sometimes showed interesting patterns of non-randomness that could lead you towards clear text, the goal of every decoding.

The key, of course, was to have a large amount of text and a fast computer to analyze billions of possible mappings. I had access to fast computers, and had been able to write some interesting software. I had published five articles on the technique which had generated some interest.

The problem with many ancient inscriptions was insufficient amount of text. Still, with the help of computers, intuition, and sheer blind luck, I've been a part of two teams that deciphered previously unreadable inscriptions.

It's an awesome feeling to bring a pre-historic culture into history. No matter how many artifacts you dig up, you can only guess what the individuals thought. Once you can read what they wrote they come alive – and that is why I am addicted to epigraphy. I finally got my second PhD, in Computer Science, three years ago, but programming always competed with my fascination with ancient languages for career choice. Epigraphy let me have both.

The department secretary called in the afternoon. Somebody from DARPA wanted to talk to me about a grant. I didn't recall applying for one there, but I suspected what it was.

I called them back. Sure enough, they were interested in cryptographic applications.

"I am Major Donald Murphy. Good to talk to you in person, Dr. Acker. I would like to follow up on a project you did last March."

"Thanks. Please call me Mitch. I will skip the usual jokes about your laws, Major."

"Appreciate that; I think I've heard all of them. Mitch, we have another project we'd like to fund, but there are certain sensitive data that would need to be analyzed. While they are not classified, we would feel more comfortable if you had a security clearance."

"I am afraid I will have to decline. I will be happy to work on anything less sensitive – as long as I am qualified, of course. But I have no wish for a clearance."

"Mitch, we value your insight and the previous work you did. Sorry to sound so melodramatic, but we think it would be in the nation's best interest if you had an unrestricted access to this data. I ask you to reconsider. We will make sure the process is as painless as possible."

"No thank you. I appreciate your compliments, but I still decline."

"If you do not have a clearance, Mitch, you may not be able to be the lead on those projects. Considering your past work, it would be a shame for you to take a secondary role."

That didn't get him anywhere. The major took a different tack.

"Why is that, Dr. Acker? Do you think there is something in your background that would cause a problem in getting a clearance?"

So, I was no longer "Mitch." Oh well.

"Not at all, Major. Aside from the process itself, there are a number of restrictions on people with clearances. I hope you realize that I am an anthropologist; a job that requires travel to locations probably not considered secure. In deciphering inscriptions, I have to communicate with scientists across the world, including countries that are not our allies. I am as patriotic as any American, but I consider my mission is to shed light on our ancestors; not on our current enemies."

After a few more minutes on the phone, Major Murphy finally rang off. We've had two government grants due largely to the success of my algorithms in decryption; and I hoped we would see more. However, I was sure NSA had plenty of experts who could apply those methods to their secret data – they were all available in published form.

The first thing I wanted to do was to enter the text into a computer. I started off by scanning it. I did this at home – I generally did all my early research at home, and with the tints and all, this was a very personal project.

Scanning the bound pages was easy enough, if tedious. As a computer geek, with the additional excuse of using the equipment for work, I had good scanners and lots of storage on my home network.

Of course there was no off-the-shelf OCR program that could interpret them. I would either have to do it by hand, or create my own patterns for character recognition; once I figured them out myself. Somewhere in my files, I had a neural network that I wrote that was pretty good, once trained.

The characters seemed almost geometric. I've noticed that while very few seemed to repeat, they shared many of the same features. I focused on the differences between the most similar characters.

It seemed to be that they were combinations of largely standard features. Not unprecedented; Hangul, the Korean alphabet, was largely created to simplify a more complex Chinese alphabet by using relatively simple building blocks for creating characters. More recently, a programming language that used overstriking — combining two or more characters in a single space – was invented for mathematical problems. The language was interesting, but never gained much popularity.