Of course, this isn't a problem specifically for macro evolutionists (are micro evolutionists really short scientists?) or even all atheists. It is a question, though, that physicalism — the position that phenomena can be explained by non-teleological physics — needs to answer. Happily, the answer is rather straightforward. (It's also worth noting that this question is philosophical, not scientific; it has nothing whatsoever to do with evolution or physical biogenesis.)
We might just as well ask: How do we get a galaxy from stars, none of which is a galaxy? How do we get a rock from atoms, none of which is a rock? How do we get macroscopic temperature, pressure or entropy from gas molecules, none of which have these properties? The Moon isn't an orbit, but if we take away the Moon, where's the Moon's orbit? How do we get Bibles from books, books from chapters, chapters from paragraphs, paragraphs from sentences, sentences from words and words from letters? How do we get meaning from marks on paper or sounds in air, none of which are intrinsically meaningful?
The obvious answer is that "life" is an abstract, emergent property. It's a property our minds ascribe to aggregations of matter that do particular things, things like move around, reproduce, use energy in a particular way, etc. There is no such thing as "life" as an independent, intrinsic, fundamental property: There is only the abstract property of "being alive" that we ascribe to aggregations of matter.
Abstract, emergent properties stand in contrast to concrete, intrinsic properties, such as mass* or extension in space*. Concrete, intrinsic properties are irreducible and additive: We cannot take away the mass of an atom without taking away the atom itself and the mass of an aggregate of atoms is the sum of the mass of the individual atoms**.
*Yes, I'm oversimplifying somewhat.
The concept of abstract, emergent properties poses some interesting philosophical questions, but it's very difficult to simply dispose of the concept and speak exclusively about only concrete, intrinsic properties (and perhaps simple relative properties such as relative velocity). Talking about abstract, emergent properties is just too valuable and useful a technique to dispose of it just because it gives an answer you don't like.
(Another way of looking at the issue is that, in some sense, life is matter in motion, and motion is a concrete, intrinsic property of matter.)
Commenter ubercheesehead had an interesting take on the issue. He abandoned the notion that emergent properties were impossible, and instead argued that emergent properties were evidence of intelligent design. The argument, I imagine, would go something like this:
- Not counting terrestrial organisms, all of the objects we can observe that have complicated emergent properties are artifacts of (human) intelligent design.
- Terrestrial organisms have complicated emergent properties
- Therefore terrestrial organisms are the product of intelligent design.
It's an evidentiary argument (complicated emergent properties are held as evidence for intelligent design), but evidentiary arguments have to account for all the evidence; it's not enough to just cherry-pick the evidence that supports your case. And, of course, there are many dissimilarities between terrestrial organisms and human artifacts: Human artifacts show lateral transfer, lack persistent trivial functional flaws, change dramatically over short periods of time, and operate cooperatively, not competitively. Other than having complex emergent properties, terrestrial life forms lack few similarities, especially features of human artifacts that we can identify as entailing from their intelligent origin.
More importantly, this sort of argument fundamentally misuses the evidentiary mode of argumentation. Showing the misuse is subtle and complicated, so I'll address it in my next post.