“Mutations that are grammatical,” she said, “physiological statements that don’t violate any important structural rules in an organism.”

Miller smiled beatifically, then held his knee and began rocking gently back and forth. His large square cranium glinted as it caught the reddish gleam of an overhead light. He was thoroughly enjoying himself.

“Where would the evolutionary information be stored — throughout the genome, holographically, in different parts in different individuals, or just in germ-line cells, or…elsewhere?”

“Tags stored in a set-aside section of the genome in each individual,” Kaye said, and then bit her tongue. Miller — and Saul, for that matter — regarded an idea as a kind of food that needed to be thoroughly shared and chewed over before it could be useful. Kaye preferred certainties before she spoke. She searched for an immediate example. “Like heat-shock response in bacteria, or single-generation climate adaptation in fruit flies.”

“But a human set-aside has to be huge. We’re so much more complex than fruit flies,” Miller said. “Have we found it already, but just don’t know what it is?”

Kaye touched Saul’s arm, urging caution. They had a reputation now for riding a certain wave, and even with an old-guard scientist like Miller, a gadfly with sufficient accomplishments under his belt for a dozen careers, she felt nervous giving away their most recent thinking. It could get around: Kaye Long says such and such…

“Nobody’s found it yet,” Kaye said.

“Oh?” Miller said, searching her face with a critical gaze. She felt like a deer frozen in headlights.

Miller shrugged. “Maybe not. My guess is, it’s expressed only in germ-line cells. Sex cells. Haploid to haploid. It doesn’t get expressed, it doesn’t start work unless there’s confirmation from other individuals. Pheromones. Eye contact, maybe.”

“We think otherwise,” Kaye said. “We think the set-aside will only carry instructions for the small alterations that lead to a new species. The rest of the details remain encoded in the genome, standard instructions for everything below that level…Probably working as well for chimpanzees as for us.”

Miller frowned, stopped rocking. “I have to let that run around in my head for a minute.” He glanced up at the dark ceiling. “Makes sense. Protect the design that you know works, at a minimum. So will these subtle changes carried in the set-aside express as units, do you think,” Miller said, “one change at a time?”

“We don’t know,” Saul said. He folded his napkin beside his plate and thumped it with his hand. “And that’s all we’re going to tell you, Drew.”

Miller smiled broadly. “Jay Niles has been talking with me. He thinks punctuated equilibrium is on a roll, and he thinks it’s a systems problem, a network problem. Selective neural network intelligence at work. I’ve never much trusted talk about neural networks. Just a way of clouding the issue, of not describing what you need to describe.” With complete lack of guile, Miller added, “I think I can help, if you want me to.”

“Thanks, Drew. We might call on you,” Kaye said, “but for right now, we’d like to have our own fun.”

Miller shrugged expressively, tipped his finger to his forehead, and walked back to the other end of the table, where he picked up another breadstick and began another conversation.

On the plane to La Guardia, Saul slumped in his seat. “Drew has no idea, no idea.”

Kaye looked up from the airplane copy of Threads.

“About what?” Kaye asked. “He seemed pretty on track to me.”

“If you or I or anybody in biology was to talk about any kind of intelligence behind evolution…”

“Oh,” Kaye said. She gave a delicate shudder. “The old spooky vitalism.”

“When Drew talks about intelligence or mind, he doesn’t mean conscious thought, of course.”

“No?” Kaye said, deliciously tired, full of pasta. She pushed the magazine into the pouch under the tray table and leaned her seat back. “What does he mean?”

“You’ve already thought about ecological networks.”

“Not my most original work,” Kaye said. “And what does it let us predict?”

“Maybe nothing,” Saul said. “But it orders my thinking in useful ways. Nodes or neurons in a network leading to neural net patterns, feeding back to the nodes the results of any network activity, leading to increased efficiencies for every node and for the network in particular.”

“That’s certainly clear enough,” Kaye said, making a sour face.

Saul wagged his head from side to side, acknowledging her criticism. “You’re smarter than I’ll ever be, Kaye Lang,” Saul said. She watched him closely, and saw only what she admired in Saul. The ideas had taken hold of him; he was not interested in attribution, merely in seeing a new truth. Her eyes misted, and she remembered with an almost painful intensity the emotions Saul had aroused in their first year together. Goading her, encouraging her, driving her nuts until she spoke clearly and understood the full arc of an idea, a hypothesis. “Make it clear, Kaye. That’s what you’re good at.”

“Well…” Kaye frowned. “That’s the way the human brain works, or a species, or an ecosystem, for that matter. And it’s also the most basic definition of thought. Neurons exchange lots of signals. The signals can add or subtract from each other, neutralize or cooperate to reach a decision. They follow the basic actions of all nature: cooperation and competition: symbiosis, parasitism, predation. Nerve cells are nodes in the brain, and genes are nodes in the genome, competing and cooperating to be reproduced in the next generation. Individuals are nodes in a species, and species are nodes in an ecosystem.”

Saul scratched his cheek and looked at her proudly.

Kaye waggled her finger in warning. “The Creationists will pop out of the woodwork and crow that we’re finally talking about God.”

“We all have our burdens.” Saul sighed.

“Miller talked about SHEVA closing the feedback loop for individual organisms — that is, individual human beings. That would make SHEVA a neurotransmitter of sorts,” Kaye said, mulling this over.

Saul pushed closer to her, his hands working to describe volumes of ideas. “Let’s get specific. Humans cooperate for advantage, forming a society. They communicate sexually, chemically, but also socially — through speech, writing, culture. Molecules and memes. We know that scent molecules, pheromones, affect behavior; females in groups come into estrus together. Men avoid chairs where other men have sat; women are attracted to those same chairs. We’re just refining the kinds of signals that can be sent, what kinds of messages, and what can carry the messages. Now we suspect that our bodies exchange endogenous virus, just as bacteria do. Is it really all that startling?”

Kaye had not told Saul about her conversation with Judith. She did not want to take the edge off their fun just yet, especially with so little actually known, but it would have to happen soon. She sat up. “What if SHEVA has multiple purposes,” she suggested. “Could it also have bad side effects?”

“Everything in nature can go wrong,” Saul said.

“What if it actually has gone wrong? What if it’s been expressed in error, has completely lost its original purpose and just makes us sick?”

“Not impossible,” Saul said in a way that suggested polite lack of interest. His mind was still on evolution. “I really think we should work this over in the next week and put together another paper. We have the material almost ready — we could cover all the speculative bases, bring in some of the folks in Cold Spring Harbor and Santa Barbara…Maybe even Miller. You just don’t turn down an offer from someone like Drew. We should talk to Jay Niles, too. Get a real firm base laid down. Shall we go ahead, put our money on the table, tackle evolution?”

In truth, this possibility scared Kaye. It seemed very dangerous, and she wanted to give Judith more time to learn what SHEVA could do. More to the point, it had no connection with their core business of finding new antibiotics.