I have previously made clear where I stand in that debate, and for what reasons, but at the same time I do have some reservations with a paper in the Journal of Human Evolution that applies cladistic analysis (also known as phylogenetics) to determine where in hominin evolution H. floresiensis fits in.
The authors, Argue et al., examined multiple morphological features of bones of H. floresiensis, H. ergaster, H. erectus, H. habilis, H. rudolfensis, A. africanus, A. afarensis, H. rhodesiensis, H. georgicus, and H. sapiens in order to compare them and construct cladograms (evolutionary or phylogenetic trees showing ancestral relationships). Because some scientists suggest that the smaller brain case could have been caused by microencephaly, they excluded cranial capacity from the comparisons.
The result is that two trees are the most parsimonious, meaning that they are the shortest of all the trees they examined. In other words, under the assumption that the fewer morphological changes between the species is the better model for the evolutionary relationship, these two trees are the most likely of all.
The two most parsimonious cladograms, with H. floresiensis branching off just after and just before H. habilis, respectively. Dmanisi is also known as Homo georgicus.
So the conclusion the authors make seems fair enough: H. floresiensis branched off either just after or just before H. habilis.
Skipping over the problems I have with the assumption of parsimony, does anyone else notice something really fishy so far?
The fishy part is that in order to construct cladograms like the ones above, the authors assumed that H. floresiensis and H. sapiens are different species. Once that is done, no other conclusion can be reached. It is perhaps interesting by itself to see what the evolutionary relationship is between them, assuming that they have one, but it can't address the question of whether they are different species or not. However, Argue et al. state clearly in their introduction that this is their goal:
Alternative interpretations include the possibility that the Liang Bua fossils represent a new hominin species, H. floresiensis (...), and that the holotype specimen, LB1, was a modern human, possibly afflicted with a pathological condition (...). These conflicting hypotheses are based on comparative analyses of the morphology of the bones with both archaic and modern Homo, typically using statistical methods to compare the Liang Bua bones with those of other hominins.Again, I do not see how using cladistic analysis can be used to resolve this question of whether H. floresiensis was a human with a pathological condition or a separate species. In fact, the same issue exists between H. georgicus (Dmanisi) and H. erectus. H. georgicus is now thought to represent an early stage before H. erectus, rather than being a separate species. Yet, here they are assumed to be different species.
The morphological and morphometric analyses have contributed much to the debate about H. floresiensis, but have not conclusively resolved the controversy about the position of the species in human evolution. We, therefore, use a different tool, cladistic analysis, which has not yet been applied to resolving this problem.
From their conclusion:
Based on rigorous cladistic analyses, we propose that H. floresiensis evolved in the Late Pliocene or Early Pleistocene. The first of our two equally parsimonious trees suggests that H. floresiensis branched after H. rudolfensis (represented by KNM-ER 1470) but prior to the divergence of H. habilis (represented by KNM-ER 1813 and OH 24). Alternatively, our results are equally supportive of H. floresiensis branching after the emergence of H. habilis.Agreed. I too believe H. floresiensis to be a new species, and given that, I can accept the cladistic analysis (again, ignoring my issues with parsimony in cladistics).
Our results sustain H. floresiensis as a new species (Brown et al., 2004; Morwood et al., 2005) and favor the hypothesis that H. floresiensis descended from an early species of Homo (Falk et al., 2005; Argue et al., 2006; Larson et al., 2007; Tocheri et al., 2007).In their earlier paper (Argue et al., 2006) the morphological data were used to support this conclusion, but, again, the results in the present paper does not "sustain H. floresiensis as a new species (...) and favor the hypothesis that H. floresiensis descended from an early species of Homo," when it is assumed that they are separate species.
Argue D, Morwood M, Sutikna T, Jatmiko, & Saptomo W (2009). Homo floresiensis: A cladistic analysis. Journal of human evolution PMID:19628252