Faced with the question ‘what is life?’ many scientists, and some philosophers, advance definitions of life. Defining life is especially popular among astrobiologists, many of whom are convinced that one cannot successfully search for truly novel forms of microbial life without a definition of life: How else will one recognize it if one encounters it? The extensive discussion of definitions of life in Part 2 (“Definition and nature of life”) of the CRC Handbook of Astrobiology (Kolb 2018) provides a salient illustration of this attitude. Along the same lines, a recent version of the NASA Astrobiology Strategy (Hays 2015) contains a large section devoted to “Key Research Questions for Defining Life” (p. 145).1 This chapter and the next explain why the scientific project of defining life is mistaken. Life is not the sort of thing that can be successfully defined. In truth, a definition of life is more likely to hinder than facilitate the discovery of novel forms of life.

There are universal theories in physics and chemistry but no universal theories in biology. The failure of biologists to come up with such a theory is not due to a lack of effort. Philosophers and scientists have struggled to formulate universal principles of life since at least the time of Newton. This chapter traces the history of these efforts back to their roots in the work of the ancient Greek philosopher Aristotle. Aristotle’s influence can be seen today in the view, which dominates contemporary biological thought about the nature and origin(s) of life, that the following abstract functional characteristics are basic to life: (1) the capacity to self-organize and maintain self-organization for an extended period of time against both external and internal perturbations and (2) the capacity to reproduce and (in light of Darwin’s theory of evolution) transmit to progeny adaptive characteristics. For the sake of simplicity, I refer to the former as “O” and to the latter as “R” throughout this chapter. As Section 1.2 discusses, the conceptual parallels between O and R and Aristotle’s ideas about life are remarkably close. He identified “nutrition” and “reproduction” as the basic functions of life and debated (as do so many contemporary researchers) which is more basic. Aristotle also bequeathed to biology the thorny problem of teleology – the notion that the allegedly basic functions of life (in their contemporary guise, metabolism and genetic-based reproduction) require a strange (to the modern scientific mind) form of causation that is intrinsically directed at achieving a future goal. As Aristotle argued, living things are not just fed, they feed themselves, and they are not just copied, they reproduce themselves. Characteristic O reflects this view in explicitly referring to the idea of self-organization. Similarly, characteristic R implicitly assumes that organisms contain an internal principle for generating organisms resembling themselves; external processes do not (like a 3D printer) duplicate them.

The most significant challenge facing the pursuit of a universal theory of life is the infamous “N = 1 problem.” In the late twentieth century biologists made an astonishing discovery. Life as we know it on Earth today descends from a last universal common ancestor (LUCA), and hence represents a single example of life. Logically speaking, one cannot safely generalize to all of life, wherever and whenever it may be found, on the basis of a single example. As Section 5.2 explains, the N = 1 problem of biology is not just a pernickety logical point. There are compelling scientific reasons for worrying that our sample of one may be unrepresentative of life. Biochemists and molecular biologists have established that life could differ from familiar Earth life in significant ways at the molecular and biochemical levels. In addition, astrobiologists have explored how the basic functions of familiar life (metabolism and genetic-based reproduction) might be realized by molecular compounds based on elements other than carbon under chemical and physical conditions differing from those thought to have been present on early Earth.

This book focuses on the search for a universal theory of life. It is concerned with the history of attempts to develop such a theory, diagnosing why these efforts have thus far been unsuccessful, and determining what is required to forge ahead and successfully pursue such a theory. It is of course possible that the diverse phenomena of life lack an objective natural unity, and hence that no such theory will ever be forthcoming. Indeed, this view has become popular among some biologists and many philosophers of biology. One of the central themes of the book is that skepticism about the prospects of universal biology is not only very premature but also potentially self-fulfilling: One does not want to short-circuit the potentially successful pursuit of universal biology by rejecting it out of hand.

This chapter explores the possibility of a shadow biosphere, that is, a form of microbial Earth life descended from an alternative abiogenesis.1 It is widely assumed that all life on Earth shares a common origin. Yet there is surprisingly little theoretical or empirical support for this belief, although it is true that all known life is so related. As Section 9.2 explains, the possibility that more than one form of life arose on Earth is consistent with (i) prevailing models of the origin of life (the RNA and SM (Small Molecule) Worlds, discussed in Section 5.4) and (ii) our current understanding of molecular biology and geochemical conditions on the early Earth. While the possibility that our planet hosted more than one abiogenesis is often conceded, many scientists nonetheless insist that any descendants would have been eliminated long ago by our microbial ancestors in a Darwinian competition for vital resources. As we shall see, this theoretical argument is undermined by what has been learned in recent years about the structure and dynamics of microbial communities.

What factors impede the development of successful scientific theories? How can the development of such theories be facilitated? These questions arise independently of which conception of scientific theory one endorses. They are especially important for biology since we currently lack a scientifically fruitful, universal theory of life; as many biologists are fond of admonishing, “give me a general principle of biology and I’ll find an exception.” At best (assuming that life has a universal nature, which is not certain) we are still in the earliest stages of formulating such a theory. For as the next chapter (Chapter 5) explains, recent advances in biochemistry and molecular biology have established that familiar Earth life provides just a single example of life. Biologists have also discovered that complex multicellular eukaryotes are highly specialized, biologically fragile, latecomers to our planet. Yet (as discussed in Chapter 1) the latter, especially animals and plants, have served as prototypes for biology since the time of Aristotle. In a nutshell, in addition to generalizing on the basis of a single example of life, we have been seeking universal principles for life from an unrepresentative subsample of it.

Some artificial life (ALife) researchers contend that we are on the verge of creating novel life forms either in the form of information structures in a computer (soft ALife), robots made of plastic and metal (hard ALife), or synthetic organisms composed of unnatural biomolecules (wet ALife or synthetic biology). In the words of computer scientist Chris Langton, a founder of ALife, “Artificial life can contribute to theoretical biology by locating life-as-we-know-it within the larger picture of life-as-it-could-be” (Langton 1989, 1). Similarly, in a discussion of synthetic life, Mark Bedau and colleagues note that “[o]ne of artificial life’s key goals is constructing a life form in the laboratory from scratch” (Bedau et al. 2000, 365). This chapter evaluates whether ALife research can live up to its hype and deliver truly novel forms of life. As will become apparent, the inventions of ALife are very closely based on characteristics of familiar Earth life. In addition, they tend to fall outside of the Goldilocks level of abstraction (Section 4.3) in being either too abstract (soft and hard ALife) or too concrete (wet ALife).

Not everyone who advances a so-called definition of life has in mind the traditional notion of definition. This is especially true of scientists. Biochemist Steve Benner (2010), for instance, contends that definitions encapsulate theories, and speaks of the need for formulating a “definition-theory of life” (p. 1022). But it is also true of some philosophers who are well aware of the limitations of traditional definitions. As an illustration, Mark Bedau (1998) presents a “definition” (his term) for life and characterizes it as the “general form of my theory of life” (p. 128). Definitions of this sort are nonstandard in the sense that their authority does not derive from analysis of human concepts or alternatively mere stipulations of meaning. Their acceptability depends upon successful empirical investigations. Nonstandard definitions nonetheless resemble traditional definitions structurally insofar as they supply necessary and sufficient conditions (identifying descriptions) for membership in a presumed natural kind. Some recent proposals for “defining” life, briefly discussed in Section 3.5, are even more radical, rejecting the received view that a central function of definition is classification; on such a proposal, a definition of life need not even provide necessary and sufficient conditions for life.

When most people think about extraterrestrial life they envision intelligent, often technologically advanced, humanoid creatures, such as the Prawns (District Nine) and the Na’vi (Avatar), robots, such as Gort (The Day the Earth Stood Still), and the hive-like Borg (Star Trek). Most astrobiologists, however, are not looking for intelligent life. They are looking for bacteria-like organisms. For as discussed in Section 5.2, microbial life is almost certainly far more common in the universe than complex multicellular organisms, let alone intelligent, technologically sophisticated creatures. Because they are so tiny, detecting and identifying extraterrestrial microbes is especially challenging.