Wolfram Computation Meets Knowledge

Wolfram Summer School


Eben Rose

Summer School

Class of 2013


I am working from my home in Maine to finish writing my PhD dissertation through the Department of Geology and Geophysics at Yale University. I am a field geologist by training who had previously tackled a large-scale project dealing with Cambrian strata (~513 million years old) of the classic exposures of the Grand Canyon, Arizona as my MS thesis. I was subsequently invited to participate in a PhD project dealing with much older strata-the oldest preserved strata on Earth at >3.0 billion years—in southern Africa and northwestern Australia. I spent several field seasons in these remote localities and have ever since been steeped in the burgeoning literature surrounding the emergence of life on Earth and the traces this has left in these ancient strata.

Traces of the earliest (bacterial) life on Earth are not so straightforward to interpret. Such traces are indirect and are highly mediated by theory, i.e., about what we expect to see that goes beyond the data. Bacteria-shaped blebs and geochemical proxies for bacterial metabolism, for example, can be attributed to a range of sources and origins, including post-burial contamination. Many such traces have been held up as “direct” empirical evidence for the earliest life, only to be shown to be something else a few years later. Ties to the equally burgeoning genre of astrobiology has pushed the interpretation of these enigmatic geological and geochemical traces toward being bacterial “life-as-we-know-it”. But I have proposed an alternative interpretation for my thesis: that these ancient strata preserve traces of a pre-biotic Earth with variously evolved and functional—but not yet bacterial—organic matter. Geologists are not trained to consider what traces such pre-biotic organic matter may leave in the rock record, and this gap of insight contributes to our tendency to assume that such traces are bacterial (in a degraded form of preservation) in ways that that are guided by analogical reasoning that compares the modern to the ancient.

Put another way, I propose that the empirical traces in this most ancient of Earth’s stratal record can support either theory equally well: that (A) they are degraded bacterial traces or that (B) they are traces of a pre-or proto-bacterial world we do not yet fully understand. Without a clear understanding of (B), we have no real basis upon which to reject (B) in favor of (A).

As it stands, geology’s contribution to the genre of origin of life studies has witnessed troubling recurrence of high-profile empirical “false-positives”. This raises questions of what constitutes explanatory success and has directed my attention toward the limitations and shortcomings of abductive reasoning within the geohistorical sciences. This project has demanded consideration of how theory meshes with empirical observation generally, and how both theory and empirical observation interact with non-empirical factors, including social and career-related motivational factors, in constructing scientific claims of knowledge about the origin of life.

Project: 3D DLA/CA Model for Formation of Conical Stromatolites

Empirical knowledge of the >2.5 billion-year-old age of bacterial life on Earth derives from three categories of traces that are examined in the stratigraphic rock record. These categories include (1) cell-like microscopic textures or traces that are/have been interpreted as fossilized bacteria; (2) various chemical traces, such as stable isotope analyses (mostly of 13C/14C and S 32, 33 ,34, and 36), that have been variously interpreted as metabolic signatures; and (3) meso-scale (~10 cm) layered sedimentary structures known as “stromatolites”. Each of these categories has limitations and caveats in how the link is made between observation of individual traces and theory that is brought to those observations.

My project focuses on stromatolites and is a followup of a 2006 publication by some former colleagues of mine (see reference below). Using their yet-unpublished 3D version of a combined DLA (diffusion-limited aggregation)/CA computer program (written in C++), an almost complete range of known stromatolite morphologies can be generated, from ramified and dendritic tree-like structures to bulbous and low-relief dome-shaped structures. The notable exception to this is that the oldest stromatolites at ~3.5 billion years old are of a shape that resembles stacked cones, and this conical shape is not modeled well by their 3D DLA/CA model. A key to understanding this unique cone-like structure is that, in addition to being layered, the natural occurrences of these very ancient stromatolites also reveal a fabric of original (now minerally replaced), 10-cm scale crystals growing radially from the center and base of the cones. A reformulated or modified model that considers the interfering influence of the growth of these crystals may account for their unique shape.

Although many geologists believe that morphology is diagnostic of life, the overall issue regarding interpretive strength of stromatolites as the oldest sign of life depends only on how we interpret the DLA parameter of attraction distance of fallout particles. This parameter relates to the adhesive property between the particle and the growth surface. In nature, this adhesive property is best understood in terms of biologically-generated organic matter (i.e., bacterially-produced exopolymeric substances). But “sticky” organic polymers may have an older history on Earth than bacteria, and so the link between these stromatolites and bacterial life is not as strong as is commonly asserted. By constraining the parameters used in models such as this DLA/CA model, it becomes easier to articulate these important distinctions.

Favorite Four-Color, Four State Turing Machine

Rule 5917190646080484251743