In my own attempts to seek out protistan group relationships in the 1970s, the conventional categories and boundaries were disregarded. This was not difficult, because I worked with natural microplankton communities in which the similarities between various groups of protists, other than the presence or absence of plastids, were more obvious than their differences. Instead, the form category ‘flagellates' was used (Taylor, 1976), or the now frowned-upon ‘lower eukaryotes' (Taylor, 1978). It had become evident through these and earlier independent efforts that protists were a diverse radiation of early diverging eukaryotes in which multicellularity had evolved repeatedly, independently: many groups had remained unicellular, but others had given rise to multicellular photosynthetic and non-photosynthetic groups including the plants and animals. Defining a formal group ‘Protista’, using Haeckel's unicellularity criterion, would orphan several multicellular groups not closely related to animals and plants, such as the ‘brown algae’, and raise problems in groups that have both. Such a kingdom would thus be a ‘grade’ or level-of-organization, which is not acceptable in formal classification among phylogenetic systematicists, since the group Protista would not include all its descendants, i.e. was paraphyletic. This is also true of the group Protoctista, which includes some multicellular descendants but not all. The avoidance of paraphyly, while theoretically laudable, seems almost impossible to reconcile with classical systematics. Alternatively, according each major clade kingdom status leads to taxon inflation to an extreme degree. The Five Kingdom/Two Superkingdom approach is intuitively appealing in its simplicity of concept, but is weak in its theoretical underpinnings and needs to be reconciled with the Three Domain (Bacteria, Archaea, Eukaryota) view arising from SSU rDNA trees. The Kingdom Protoctista has been strongly advocated by Margulis in the Handbook (Margulis et al., 1990) and as one of the ‘five kingdoms' that she has popularized significantly in recent years (Margulis & Schwartz, 1998). Although fuzzy on detailed relationships, there is no doubt that these works have substantially opened the minds of educators to the protistan alternative.
Given the state of knowledge prevailing in the 1970s and 1980s, a detailed new classification beyond that of Copeland (1956) would have had to undergo repeated conceptual and nomenclatural changes. Others have accepted the challenge, notably Cavalier-Smith (1981, 1993, 1998; and many other contributions up to this volume). This iterative approach should eventually lead to a more robust classification but, unfortunately, it generates new names and differing uses of old names as it goes. Corliss (1972) appealed for common sense and courtesy in such matters to avoid the proliferation of new higher taxa. The whole systematic question has been reviewed in detail by several authors, including Möhn (1984), Cavalier-Smith (1998), Corliss (1998a, 2002) and Patterson (1999, 2000).
Another approach, currently in vogue, is to depict the relationships of many of the definable 75 or so protist lineages as a ‘comb’, in which most have no sister groups (e.g. Patterson, 1999, 2000). However, the aggregation of groups according to mitochondrial crista types, as Patterson (2000) has noted, does seem to be remarkably concordant with several molecular datasets, including rDNA and mitochondrial DNA.
The literal meaning of the formal names of protist groups currently in use has been commented on elsewhere (Taylor, 1999). I regret the continuation of anachronistic and inappropriate group-endings such as -zoa, -phyta and -mychota, except where they are literally apt. Also, the use of ‘Protozoa’ to include many more groups than usual, including some that are poly- or paraphyletic, seems unfortunate and likely to mislead the unknowing or unwary [see Patterson (2000) for a more detailed criticism and Corliss (2002), which includes a brief rebuttal].