The assumption that all life on Earth today shares the same basic molecular architecture and biochemistry is part of the paradigm of modern biology. This paper argues that there is little theoretical or empirical support for this widely held assumption. Scientists know that life could have been at least modestly different at the molecular level and it is clear that alternative molecular building blocks for life were available on the early Earth. If the emergence of life is, like other natural phenomena, highly probable given the right chemical and physical conditions then it seems likely that the early Earth hosted multiple origins of life, some of which produced chemical variations on life as we know it. While these points are often conceded, it is nevertheless maintained that any primitive alternatives to familiar life would have been eliminated long ago, either amalgamated into a single form of life through lateral gene transfer (LGT) or alternatively out-competed by our putatively more evolutionarily robust form of life. Besides, the argument continues, if such life forms still existed, we surely would have encountered telling signs of them by now. These arguments do not hold up well under close scrutiny. They reflect a host of assumptions that are grounded in our experience with large multicellular organisms and, most importantly, do not apply to microbial forms of life, which cannot be easily studied without the aid of sophisticated technologies. Significantly, the most powerful molecular biology techniques available-polymerase chain reaction (PCR) amplification of rRNA genes augmented by metagenomic analysis-could not detect such microbes if they existed. Given the profound philosophical and scientific importance that such a discovery would represent, a dedicated search for 'shadow microbes' (heretofore unrecognized 'alien' forms of terran microbial life) seems in order. The best place to start such a search is with puzzling (anomalous) phenomena, such as desert varnish, that resist classification as 'biological' or 'nonbiological'.