The Red Syndrome (14 page)

"The percentage of frequency of each letter of the alphabet in an
English text."

"And if the message isn't in English? Then what?"

I had no answer. "Honestly, I didn't think about it. You're right. The
encrypted text could be in Portuguese, or Swedish, or chart znayet."

"What?"

Ahh." I smiled. "It means `only the devil knows' in Russian."

Laura laughed, exposing her perfect teeth. "I thought you were cursing.
Do you speak Russian?"

"I picked up a little from a Russian girlfriend I had some years back.
But here's what's important now: Every language follows certain linguistic patterns, the most obvious of these being letter frequency. As the
length of the text increases, so does the likelihood that it will begin to
follow typical patterns. The decoding of the encrypted text is based on
comparisons between the frequencies of characters in the encrypted text
and frequencies of characters in a plain text."

"Okay, then the statistical tools we should use are summary statistics
and sampling. I can't give you numbers off the top of my head. Only linguists could do that."

"Do you know any?"

Laura hesitated. "There's an old professor, Alexander Klebanov. He
must have retired by now, but he was always giving us problems that
included calculating probabilities of letters in random sentences."

"That's our guy," I said excitedly. "Could you call him?"

"Now?"

"Sure."

She lifted the receiver and dialed the ¢11 national directory.

Five minutes later Professor Klebanov was on my speakerphone from
his Florida home. After the absolutely necessary niceties, including three
long minutes of Laura trying to remind the professor who she was, she
finally posed the question.

"I need to know in what language you assume the message was
written," he answered in a slight accent. "Asking the right question is
almost as important as getting the right answers." I imagined him saying
that before a class full of college students.

"We don't know, but let's assume it was English."

"Next, I need to know if it was written in American modern English
or, say, Shakespearean English."

"Same answer, no idea. If it's English, I'd suspect the language used was
modern to be able to include contemporary terms such as bank or currency," she said.

Obviously, she was right: Shakespeare didn't receive royalties through
direct wire transfers into his bank account.

Laura was smart. I liked smart women.

"I think there's simple computer software that analyzes frequency of
letters," said Professor Klebanov.

"Does it work in reverse as well?"

"Meaning?"

"Can it decrypt?"

"Maybe," he responded. "But why don't you fax me the document you
want to break, and I'll try doing it here. I have always liked these mental
games.

"Sure," said Laura.

"Sorry, we can't," I interjected. "Thank you for your kind offer, but this
is government material. I hope you understand."

"Of course," said Professor Klebanov. "Well, in that case, I have software in my home computer that does the calculations, so if you give me
a random sentence, I could analyze the letters' frequency. That would give
you the probabilities you wanted. But to be more accurate I suggest you
use the table calculated by Beker and Piper."

"Do you happen to have a copy?"

"I have a book with the table in it. I can fax it to you. I can't tear out
the page, so I hope you won't mind me copying it longhand. Do you have
a fax machine available?"

"Kind of. I have a fax card in my home computer."

"That'll do it. Give me the number."

I gave it to him.

Laura thanked Professor Klebanov and hung up.

Twenty minutes later a handwritten message came in by fax.

"That's great," I said enthusiastically. "Now let's see if we can get anything out of the text."

Laura was cautious. "Remember that these numbers are based on sample
writing. Frequency of certain letters will change from one text to another,
but within the margin of error."

I moved to the couch next to Laura, who stretched out her legs, leaving
a very narrow space for me, almost touching her body. "I'm comfortable
here," she said. "I hope you don't mind."

"Not at all." Her head was close to mine and I could smell her. Although
she told me that she'd gone out to smoke a cigarette, there was no tobacco
scent; only a gentle perfume.

I took the top message of the three I received from Detective Mahoney,
printed another copy of Professor Klebanov's table, and gave it to Laura.
I put my copies on the small coffee table next to us.

"It seems that once you break the first ten letters, the rest are like completing sentences in Wheel of Fortune," she said.

I felt pressured. I went over the encrypted text, almost praying that the
ciphered text was in English and the encryption was not deep. I gulped
from my beer and felt the tangy bitterness cool my throat.

"What letter appears in the message most frequently?" she asked.

"Let me count ... I think the letter X," I answered. "So we could
assume it was replacing A, which according to the table is the most frequently used letter in English." That came as a surprise to me, though. I
counted again, "The next most frequent letter in the ciphered text is Y"

"It could be I." She pointed at the chart, which said I is the third most
common letter in the English language.

The third most frequent letter was G. "According to the chart, I
assumed it was an S," she said.

I paused. "Something must be wrong," I had to concede. "There is no
word in English that starts with the sequence ais."

"How about aisle?"

"Okay. Any other word?" I asked.

Laura was quick to evaluate the situation. "I don't think so. If it's not
aisle, then there are at least two explanations. The first is that the ciphered
text is in a language other than English."

"And the other?"

"That the text was encrypted in a more sophisticated method than the
one we re assuming.

It was possible the work we had already put in was worthless. I took
another sip of my beer and started to think about an alternative, more
joyful manner to spend the rest of the evening with Laura. I had a sense
she thought so, too, hence her apparent unenthusiastic behavior.

Her voice brought me back to the cold reality. "I don't think we have
any alternatives at this time."

"So you want to quit?" I asked.

"I think so," she said.

"Well, let's give it a try and continue with the two assumptions, that the
text is in English and that the encryption is substituting letters."

"To make things even more complicated," she said with a grin, sensing my frustration, "the chart of frequencies is statistical, and statistics have a
margin of error. So although X, Y, and G are used most frequently, it does
not necessarily mean that they were used in that order."

"I know," I confirmed. "We simply can't be certain that X equals A, Y
equals I, and G equals S."

"True," she agreed. "So let's assume for a minute that X, Y, and G stand
for E, L, and R, because the probabilities are close and the text is relatively short."

I nodded. I saw what she was aiming at and I liked it. I continued her
train of thought. "To be more precise, we must look at the letters next to
the most frequent letters, because grammatical rules limit the probability
of a certain letter being next to another. For example, the letter Q is
almost always followed by U, but is never found next to Z, unless it's a
foreign name."

When Laura just nodded, I added, "And T never comes before or after
B, D, G, and, I think, a few other letters."

"Okay," she said, "I give up, and I'm exhausted. It's too late for me to
get a cab to go home. I'm uncomfortable to travel alone that late. Is there
a couch I could use?"

"No, I'll use the couch. You take the bedroom."

"Are you sure?"

I wasn't. In fact I hadn't changed my sheets in five days. But what the
hell. Let her smell my scent as well. I showed her my bedroom and, for a
minute, entertained the idea of suggesting I join her. But she gave me no
sign of encouragement, and a rejection at this stage would foul things up.
So the complete gentleman, a role I rarely played, I gave her one of my
extra-large T-shirts, bid her good night, and fell on the couch in the
living room. I closed my eyes and thought of Laura. She was so close, just
a few feet away. I tried to think of something less tantalizing. I rolled
from side to side but couldn't fall asleep. I got up, turned on my desk
lamp, and looked at the encrypted documents. Maybe they had used a
onetime pad and I was wasting my time. But what if they hadn't? I
decided to give it another try. Maybe Professor Klebanov had erred, or
the sampling error had twisted the results.

I decided to calculate the probability of letters that were vowels and
consonants. I picked up yesterday's newspaper from the bin, copied
twenty sentences, and counted the letters. Half an hour later using my
calculator, I had a completely different probability table. Professor Klebanov must have been looking at numbers for a different language.

I looked at the combination XX. It appeared twice, but neither Ynor G
appeared doubled anywhere, so that supported the assumption that X was
replacing E.

Two hours, half a bottle of wine, and one beer later, I was still struggling. But I'd cracked eleven letters, enough to see that I was on the right
path.

I looked at the broken words and thought, Good! The text was written
in English!

The only conclusion was that the text had been encrypted in monoalphabetical cipher. That was a relief, and a warm sense of achievement
came over me.

I put together the letters I'd identified while the still-ciphered letters
were substituted with blank spaces. I highlighted the letters I'd identified
with a yellow marker.

I read out the text softly, hoping it would make some sense.

"ma-eth reetr ans-e rs-ni nehun dredt h-usa ndusd--ar seach t-acc -unth
h-tth reeth reeze r-ne at-an -schi -eri nsa-z burgd -n-tn ameap ayeem
eeting withc -mrad et-ex chang eison march secon dtran s-ers must- ecomp -
eted be-or ethen sa-eh."

It didn't mean anything, it was crazy. A trace of despair crept over me.
And I was tired.

I had a drink of soda and got another idea. Maybe I could break the
blocks into normal-sized words of two, three, four, or five letters.
Currently the message was broken into five-letter blocks. Nothing is
really written like that.

But on a second thought, why two, three, or four letters? There are also
English words that have only one letter, I orA. I continued working on it,
trying to keep my eyes open. I was really exhausted. But I wanted to break
it more than anything else. Well, almost. Finally, I whispered it aloud:

"make three transfers of nine hundred thousand usdollars each to account
hhkt three three three zero one at bank schiller in salzburg do not name a payee
meeting with comrade to exchange is on march second transfers must be completed before that day saleh."

What the hell? This wasn't plain bankers' talk. What was this bull
about a comrade? And the signature, Saleh: an Arab or Muslim name?
The message reeked to me of conspiracy.

Ishould call Hodson, I thought, but then I looked at my watch: 3:0o A.M.
It could wait until morning.

I crawled back to my makeshift bed on the couch. I heard a beep; then
another, and another. I couldn't fall asleep with some damn electronic
gadget beeping, so I got up and looked around: my cell phone? No. The
beep came again. It seemed to come from Laura's purse. It was dark and
I didn't want to turn on the light. I put my hand into her purse searching
for it, and felt a cold metal. I turned on a desk lamp. It was Laura's service
revolver. I emptied her bag and found her phone. As I was about to turn
it off, I changed my mind. I pressed the SEND button and wrote down the
last number she'd called. Was it just idle curiosity to know whether she
had a boyfriend, or was that the persistent suspicious devil in me? I
turned off her phone and returned to bed.