A History of Entomological

Classification

Michael S. Engel1, and Niels P. Kristensen2

1Division of Entomology, Natural History Museum, and Department of Ecology &

Evolutionary Biology, University of Kansas, Lawrence, Kansas 66045;

email: [email protected]

2Natural History Museum of Denmark, DK-2100 Copenhagen , Denmark;

email: [email protected]

Keywords

Insecta, orders, phylogeny, taxonomy, entomology, systematics

Abstract

The classification of insects has attempted to most effectively communicate

information about this hyperdiverse lineage of life and, not surprisingly, has

had a considerably rich historical development. This history can be coarsely

segregated into four periods: the Pre-Linnean era, the first century spanning

Linnaeuss Systema Naturae to Darwins On the Origin of Species, the

Darwinian era up to the Cladistic Revolution, and the Hennigian era leading

to today. The major events of each of these episodes are briefly summarized

and some of the more notable researchers highlighted, along with their in-

fluence on our current understanding of insect relationships and how this is

reflected in the current classification of the Hexapoda.

INTRODUCTION

Classifications have existed as long as humanity. It can be argued that the first instinct is to classify.
We classify those things that are important to us and for which we wish to find meaning and value.
Given the overarching importance of classification to every aspect of scientific entomology, it is
therefore worthwhile to consider the origins and evolution of our current system of insects.
Naturally, much has been and can be written about the history of entomology in general. An
excellent source book is Smith et al.s (129) History of Entomology, published as a special issue of
the Annual Review of Entomology, and Wilson & Doner (142) provide a somewhat disjointed but
overall interesting account of the history of insect classification. Herein we restrict ourselves to
major works that focus on the higher-level classification of insects and make no attempt to be
exhaustive. In selecting images we chose to highlight individuals or iconography not so prominently
displayed in other entomological works, for example, attempting to avoid repetition with those in
Willmanns (141) and Grimaldi & Engels (45) historical accounts. In addition, we have given cursory
treatment to the current era of entomological classification, as a compromise had to be made
regarding what should be covered. We preferred to devote such pages to those eras which form the
foundation of present-day work rather than enumerate in detail the events surrounding us. We
therefore beg the readers forgiveness for glossing over many fascinating events and ideas,
regardless of how seemingly grand or quizzical, in the historical development of entomological
classification.

THE PRE-LINNEAN PERIOD

One of the earliest surviving classifications of the natural world is that of Aristotle (384322 BC) (6),
disciple of Plato and mentor to the young Alexander the Great. Certainly there are earlier accounts
of insects, but Aristotles Historia Animalium (6) is the earliest surviving work in which a
comprehensive organization of insects is to be found, incorporating them among a larger account
of all life (see sidebar Insects in Antiquity). It was this system, in one form or another, that would
persist for the next fifteen centuries. Naturally, Aristotle and other ancient authorsGreek, Roman,
or otherwiseproposed groups of lasting value. They identified and differentiated bees, wasps,
ants, butterflies, flies, beetles, and so forth. These systems of organization persisted for millennia
and, although gradually augmented and adapted for different purposes, changed relatively little.
Thus, while Gaius Plinius Secunduss (2379 AD) Naturalis Historia was largely independent in the
biological accounts it provided, it preserved continuity in classification (129).

INSECTS IN ANTIQUITY

Parables from the Holy Bible, such as the Septuagints Go to the bee, and learn how diligent she is,
and what a noble work she produces, whose labors kings and private men use for their use, she is
desired and honored by all, and though weak in strength she values wisdom and prevails (Proverbs,
6:8), highlight the way in which people of antiquity keenly observed insects. Insects also figure
prominently in less flattering light, such as the plagues upon Egypt (Exodus, 8:1619; 10:46, 10:12
15). Although some of the notions regarding the origins of insect species were remarkably fanciful,
such as bees emanating from the carcasses of bulls, authors of antiquity simultaneously made
amazingly accurate observations. Historia Animalium rightly grouped whales among the mammals
and provided brief accounts of the dancing of honey bees, and it must be recalled that even before
Aristotle, Anaximander had proposed a mechanism of evolution, one that included humans!

Medieval Christian scholars did their best to copy what elements of knowledge they felt most
important, and perhaps one of the most enduring works of this time was Saint Isidore of Sevilles
(ca. 560636) Etymologiae(7). This singular tome attempted to provide in one work an encyclopedic
treatment of human knowledge, including a classification and account of the natural world. The
Etymologiae included accounts of termites, silkworms, and cantharids, among others, classified as
De verminibus (vermin), while things such as bees, wasps, moths, butterflies, locusts, roaches,
cicadas, and true flies were classified as De minutis volatibus (tiny flying animals). The eleventh
century brought forth presumed lost works by authors of antiquity, as well as improvements upon
them or outright novel accounts by Arabic scholars. These discoveries opened a floodgate of
information, although it took centuries to spread across Europe. The Renaissance saw little in terms
of formal classification. Instead, a diversity of scholars sought to look at insects anew, using the new
rational tools of the era, as well as the latest in technological achievements such as optics, and
disseminated their findings using the newly developed printing press. This was an exciting time for
entomology, with magnificent works by Jan Swammerdam (16371680), Marcello Malpighi (1628
1694), Antonie van Leeuwenhoek (16321723), and others investigating in greater and more
accurate detail the anatomy of insects (129), but few considered major classificatory arrangements
across the Insecta. Of note was Ulisse Aldrovandi (15221605) of Bologna, who produced the
massive folio work De Animalibus Insectis Libri VII (3), a tome that represented the first specialized
text exclusively on entomology. Aside from classification, he also gave a summary of insect
morphology, metamorphosis, reproduction, behavior, and physiology, among other biological
annotations. Thomas Mouffets (15531604) Insectorum sive Minimorum Animalium Theatrum
(104), a work amalgamating the efforts of numerous naturalists, was a related attempt at classifying
insects and their habits but lacked the finesse and detail of Aldrovandis effort. However, many of
the terms and names used by Mouffet were subsequently adopted by Linnaeus as names for his
genera. Toward the beginning of the Enlightenment we find authors again more rigorously exploring
the taxonomy and classification of insects. John Ray (16271705) relied heavily on the anatomical
work of Swammerdam and attempted to develop a classification of insects based on their
morphology, biology, and ecology, summarized in his posthumously published Historia Insectorum
(114). Although his classification had overlapping categories, confusing to a modern reader, it was
immensely influential and many of the names he applied to groups were adopted by subsequent
authors, including Linnaeus. More remarkably, Ray brought forth a definition for species and
codified the notion that species be applied to all individuals from the same progenitor. Three other
individuals worth mentioning at this time are Maria Sibylla Merian (16471717), Rene Antoine
Ferchault de R eaumur (16831756), and August Johann R osel von Rosenhof (1705 1759).
Merian published her findings in 1705 as the sumptuously illustrated Metamorphosis Insectorum
Surinamensium (98). Reaumur was perhaps the most tireless observer of insect biology and
behavior of the time (115, 129). He openly disliked the efforts of developing classifications, instead
focusing his sharp intellect on the life histories of his subjects. Rosel von Rosenhofs engravings
and life-history accounts were subsequently bound together in four volumes as Insecten-
Belustigung (120, 129). All three would have a lasting impression on subsequent entomological
authors, providing much original information about numerous insect species for the likes of
Linnaeus and Fabricius, who would never have the chance to observe those species directly or learn
anything of their habits

THE FIRST CENTURY (17581859)

Around the middle of the eighteenth century, Carl Linnaeus (17071778) adopted uniformly his
system of binomial nomenclature and in the tenth, revised edition of his Systema Naturae put

LINNEAN INSECTA Linnaeus recognized all arthropods as his fifth class (in 1758), or class Insecta,
segregating them into a series of seven orders based on the nature (or outright absence) of their
wings. Hence, based on the primacy Linnaeus gave to wings over other characters of insect anatomy,
we arrived at the familiar (in order of their appearance in his system) Coleoptera, Hemiptera,
Lepidoptera, Neuroptera, Hymenoptera, Diptera, and Aptera. Linnaeus did not focus solely on single
characters, however, and even in earlier editions highlighted the structure of the antennae among
other traits, but had to place emphasis somewhere. The orders were grouped by number of wings
(0, 2, or 4), then those in the alae 4 category were divided into superiores versus omnes, the former
containing the Coleoptera (or crustaceae totae, fully hardened) and Hemiptera (semicrustaceae,
partially hardened) and the latter containing first the Lepidoptera (imbricatae squamis,
overlapping scales) and then those with membranous wings (membranaceae, including the
Hymenoptera and Neuroptera). Linnaeus then divided the orders into his numerous genera and
species.

forth a classification of insects (88). This system was in use in earlier editions and incorporated much
information from earlier writers such as Ray and Merian, among others, but retrospectively is taken
as the starting point for our modern system of zoological nomenclature. Insect orders were
consistently applied as early as the first edition (1735), recognizing four orders {Coleoptera,
Angioptera [later Gymnaptera in the second edition (1740)], Hemiptera, and Aptera}, but by the
time of the Fauna Svecica (1746) he had arrived at an arrangement of seven orders to which he held
for the remainder of this life (see sidebar Linnean Insecta). It is often overlooked how much
importance Linnaeus gave to the authors of antiquity, but as noted, the intellectual impact and
importance of ancient Greek and Roman texts and the improvements upon them made by the early
Arabic scholars were indeed still palpable during the Enlightenment. Following upon the comments
Linnaeus made in the second edition of Systema Naturae about the essential nature of the structure
of the mouthparts, his foremost entomology student Johan Christian Fabricius (17451808), the first
systematist considered to specialize on insects, published in 1778 his Philosophia Entomologica (40),
the first textbook of entomology. Herein he synthesized all available information on insects,
including a rational means for their description and an exposition on the primacy of the structure of
the mouthparts. Fabricius believed that mouthparts would serve to establish a classification that is
natural for genera and species (given that they are vital to the nourishment of every insect and
thereby affect their whole biology) but remain artificial at the level of orders. He felt that a natural
system for the orders was for later generations, noting in 1778 that, we are only novices in the
science (40, 129). Fabricius had already set forth his system of insects in 1775 (Systema
Entomologiae), his system of genera in 1776 (Genera Insectorum), and then numerous works on
species, culminating in individual accounts of the species of each order. Given the primacy he placed
on the mouthparts, he expanded the number of orders and altered the names of those recognized
by his mentor, ultimately segregating the worlds fauna into Eleutherata (Coleoptera), Rhyngota
(Hemiptera), Piezata (Hymenoptera), Antliata (Diptera), Glossata (Lepidoptera), Ulonata
(Orthoptera), Synistata (Neuroptera), and Odonata. Baron Charles De Geer (17201778), whose
works were modeled on those of Reaumur, was the first to provide a meager attempt at uniting
the systems of Linnaeus and Fabricius, noting the artificiality of classifying together insects with
haustellate and mandibulate mouthparts, despite similarities in their wings. He therefore attempted
a simple unification of the Linnean and Fabrician systems, using both wings and mouthparts to arrive
at a slightly expanded system of orders, giving us the Dermaptera (although his use of this name
differed greatly from our modern concept) (33). This step was the impetus for subsequent authors
to reconsider the composition of the Linnean and Fabrician orders, eventually leading Guillaume
Antoine Olivier (17561814) to propose the Orthoptera (107) and much later to the disintegration
of the Neuroptera. De Geers commentator, Andres J. Retzius (17421821), simplified the
arrangement of groups and character suites (116) but erroneously ascribed suctorial forms to De
Geers Dermaptera. Linnaeus, Fabricius, and others such as Etienne-Louis Geoffroy (17271810)
noted that while their classifications at one level or another remained artificial, with each addition
of information to the hierarchy, they were getting closer to a natural system (43). Giovanni Antonio
Scopoli (17231788), however, was the first to note that the classification should be based on the
whole structure of the insect, and not just on the wings, the mouthparts, the antennae, or the habits
of the organisms in question (123). He was unable to develop a more synthetic and natural system.
A holistic approach only materialized with the French entomological luminary Pierre Andre Latreille
(17621833), whose 1796 Precis des Caract`eres Generiques des Insectes introduced the rank of
families and a synthetic approach to the classification, utilizing multiple traits, bringing together the
works of Linnaeus, Fabricius, and others, and developing the first classification professed to be truly
natural (86). In 1796 Latreille recognized Coleoptera, Orthoptera, Hemiptera, Neuroptera,
Hymenoptera, Lepidoptera, Diptera, Suceurs, Thysanoures, and Parasites, among other arthropods,
all as belonging to Insecta. This system was regularly augmented, reflecting the latest available
information on insect diversity over the next 35 years, as well as discourse with his colleagues
(including Fabricius, who visited Paris many times over several years). Latreille and his
contemporaries in Paris, such as the famed Etienne Geoffroy Saint-Hilaire (17721844), Jean-
Baptiste Lamarck (17441829), and Georges Cuvier (17691832), were making profound strides in
detailed comparative anatomy, including the first codification of a set of criteria for the recognition
of homology, debating incipient advances toward evolutionary theory and the classification of the
natural world (4). Lamarck (83, 84) expounded the philosophical view that animals tended toward a
circular process, one in which the arrangement of organisms diverged in opposite directions for
some characters, while meeting again in another character. Such concepts were carried forward by
Johann Gotthelf Fischer von Waldheim (17711853), who used Lamarcks ideas in classifying insects
and other animals (41). In terms of entomological classifications, such circularity in natural
progression saw its greatest advocate in William Sharp MacLeay (17921865). MacLeay developed
and promoted the Quinarian system of classification, whereby all higher categories of the animal
kingdom were segregated into groups of five. These were each pictographically represented by a
series of five circles bordering each other and forming a complete ring (Figure 1a). It was his notion
that each of these most closely resembled the next bordering taxon in regard to a particular
character(s), but that this transition series eventually came back upon itself. He believed strongly
that no grouping could be considered natural if it did not tend to exhibit a circular series. He also
believed that the primary divisions of such series were in groups of 10, 5 of which were the primary
elements of the series, the others being minor (osculant) and intercalating or bridging between
those of the major series. Starting first with the classification of Scarabaeus, MacLeay eventually
developed this concept into a classification of all Animalia (91). MacLeays Horae Entomologicae; or,
Essays on the Annulose Animals was hugely influential, albeit quickly recognized to have
fundamental flaws as the numbers of species and the characters they embodied soon failed to truly
fall into nice categories of five. This did not deter others from ascertaining that the true flaw was
the reliance on five, rather than the overarching philosophy itself. Indeed, Edward Newman (1801
1876) expounded a largely identical system but instead considered the fundamental elements to
organize themselves into groups of seven (105, 106). The Septenary system had its own followers
(like Swinton) but suffered the same fate as that advocated by the Quinarians.

Figure 1
(a) William S. MacLeays Quinarian arrangement of the insect orders from the Horae Entomologicae
(91). (b) A phylogeny of insect orders as presented by Hyatt & Arms (62) in 1890, with the disc
representing the surface of the Earth and the lines of descent (as interpreted from Scudders work)
extending back into the geological past. (c) Carl J.B. Borners 1904 depiction of relationships among
the principal lineages of hexapods (12). (d ) Guy Chester Cramptons 1938 remarkably modern-
looking topology of insects (30).

At about the same time as MacLeay, William Kirby (17591850), who was principally interested in
the aculeate Hymenoptera and Coleoptera, was the first to recognize the order Strepsiptera and
later the Trichoptera. In collaboration with William Spence (17831860) he produced An
Introduction to Entomology, four volumes covering all aspects of insect science including a revised
system of the higher categories, incorporating many ideas from MacLeay, as they meshed nicely
with his idea of harmony in the orderly hand of God (69, 129). In his doctoral dissertation of 1829,
De Insectorum Systemate Naturali (18), Karl Hermann Konrad Burmeister (18071892) set forth a
brief explanation for a revised system of the insect orders, and later expanded this into his textbook
of general entomology, Handbuch der Entomologie (published in five volumes between 1832 and
1855). Burmeister greatly disliked Latreilles system and belittled him for changing it so regularly.
Instead, Burmeister emphasized the different kinds of metamorphosis he observed and, estimating
the worlds fauna to consist of about 80,000 species, segregated them into eight orders
Hemiptera, Dictyoptera, Neuroptera, Diptera, Lepidoptera, Hymenoptera, Orthoptera, and
Coleoptera. Burmeisters system still emphasized a very limited set of characters, but did make
rudimentary attempts to bring together diverse data such as metamorphosis alongside wing and
mouthpart structure (19). Perhaps the last of the great entomological systematists to propose a
classification of the orders prior to the Darwinian era was John Obadiah Westwood (18051893),
the first Hope Professor at Oxford University. His two-volume An Introduction to the Modern
Classification of Insects (1839 1840) attempted to unify the systems of Latreille and MacLeay (137).
Although not a Quinarian, Westwoods deference to MacLeay is apparent throughout his
introductory remarks regarding the distribution of insects into orders. Westwood employed the
same idea of first recognizing the older orders (those of the earlier authors), and then seeking to
classify osculant groups that bridged these grander groupings, but never restricting himself to any
predetermined numerical arrangement. Accordingly, Westwood indicated that the class of
Hexapod metamorphotic insects consisted of two descending series (subclasses): those with a
mouth of jaws (Dacnostomata) versus mouth with a sucker (Antliostomata). The former
included the series order Hymenoptera, osculant order Strepsiptera, order Coleoptera, osculant
order Euplexoptera (the earwigs), order Orthoptera, order Neuroptera, and order Trichoptera, with
Westwood hedging what might be the osculant groups representing bridges between the last three.
The Trichoptera then represented a direct passage to the arrangement of the other subclass,
although this time presented in ascending fashion as order Lepidoptera, order Homoptera, order
Heteroptera, osculant order Aphaniptera (fleas), osculant order Homaloptera (Hippoboscidae and
Nycteribiidae), and finally order Diptera. The Thysanoptera were perplexing to Westwood and he
hesitatingly considered them as osculant between Orthoptera and Neuroptera (enticingly
reminiscent of the placement of Paraneoptera as between a more basal orthopteroid stem and the
holometabolans!). Westwood never accepted evolution and was an ardent opponent of Darwinism.
His system was clearly established outside an evolutionary framework, but by describing and noting
such consistent patterns of characters, Westwood was strengthening the higher classification of
hexapods while unwittingly laying the early foundation toward a phylogenetic arrangement of the
orders.

THE DARWINIAN ERA: HAECKEL TO HENNIG Work on the higher classification of insects continued
largely unchanged following the publication of Charles R. Darwins (18091882) On the Origin of
Species in 1859. To the extent authors adopted evolutionary thinking, the innovation was evident
in rhetoric rather than methodological approach, with the emphasis continuing to be on the
perceived significance of different characters. Indeed, Darwin himself noted, Systematists will be
able to pursue their labours as at present (32),

and the enduring patterns described by taxonomists for generations were now newly enlivened by
a powerful, synergistic explanation of descent. Refinements in methods of microscopical anatomy
from the late-nineteenth century onward led to numerous excellent studies on the structure and/or
developmental origin of selected organs or body regions of many kinds of insects, but the findings
went unnoticed by several systematists, who lived in different worlds, continuing to emphasize a
limited number of characters in the exoskeleton. James Dwight Dana (18131895) emphasized the
importance of cephalization in ordering the Animal kingdom, and he emphasized organizing the
insects on the basis of the degree to which more complex biological functions were centralized to
the anterior of the species (31). Dana was not alone in this endeavor. Both Gustav Schoch (1833
1899) and Frantisek Klap alek (18631919) attempted to organize the insects largely on the basis
of the structure of the thorax, effectively emphasizing their modes of locomotion, while John
Bernhardt Smith (18581912) emphasized an amalgamation of mouthpart and thoracic characters
(142). Benjamin Cooke (18161883) and Ferdinand Anton Franz Karsch (18531936), building on
the earlier work of Newman, did the same but organized around the nature of the pupa, and Carlo
Emery (18481925) and Veit Graber (18441892) more generally organized around the
development and embryology of insects (142). Such authors, then, placed emphasis on those traits
they considered most critical in the evolutionary history of the class, or outright established what
they believed were plausible mechanisms for driving macroevolutionary patterns. As in many things,
it was the controversial Ernst Heinrich Philipp August Haeckel (18341919) who first depicted, in his
1866 Generelle Morphologie, an explicit phylogeny of the insect orders, here embedded within a
topology encompassing all Articulata (45, 141). From this time onward, the classification of the
orders was increasingly presented alongside a depiction or explanation of their lines of descent,
such as the 1876 trees of Paul Mayer (18481923), who also attempted a reconstruction of the
ancestral insect Protentomon (97). A somewhat later example was the depiction of an explicit
evolutionary tree (Figure 1b) by Alpheus Hyatt (18381902) and Jennie M. Arms (18521937), who
were also the first scientists to consider the Ephemeroptera as a distinct order (62). In terms of
serious entomological inquiry, however, it was arguably Friedrich M. Brauer (1832 1904) (16) who
made the first truly conceptual leap toward classifying insects along the lines of Darwinian evolution.
After expounding upon the general principles of organic evolution and the unifying aspect of
Darwins theory, Brauer set forth his results of applying said principles to the subject of insect
diversity, effectively distinguishing between primarily wingless taxa, his Apterygogenea (Collembola
and Thysanura), and the secondarily wingless taxa that together with the winged forms constituted
his Pterygogenea, which were derived from the apterygogeneans (16). He thereby emphasized
suggestions made earlier by himself (15) and Lubbock (90), while contradicting Mayer, who had
considered the primarily wingless hexapods coordinate with the winged insects rather than
ancestral to them. Moreover, Brauer segregated the pterygotes into two broad categories,
Polynephria and Oligonephria along one axis and their development along another axis (Ametabola
+ Hemimetabola versus Metabola). Brauers system had overlapping means of segregating the
groups. If followed completely, it left him with an arrangement of 16 orders that supported earlier
attempts to divide the artificial Hemiptera (all sucking insects) and Neuroptera of Linnaeus, and
the orthopteroids, which were focused on first by De Geer, Olivier, Latreille, Leach, and Burmeister.
Difficulties remained, such as the Corrodentia of earlier authors, which for Brauer consisted of
termites, bark lice, and true lice. Brauer also seems to have been the first to segregate what we
today recognize as the Mecoptera into a distinct order, the Panorpatae. This system was supported,
although not necessarily agreed upon, by John Henry Comstock (18491931) in his influential
textbooks on entomology. Although bristling at the aforementioned composition of Corrodentia,
Comstock (26) accepted the arrangement and even further united them in a retrograde system with
mayflies, dragonflies and damselflies, and stoneflies as the Pseudoneuroptera, an antique concept
put forth by Erichson (39). Comstock did, however, over the years recognize the artificiality of the
grouping, eventually recognizing the Isoptera, Ephemeroptera (as Ephemerida), Odonata,
Plecoptera, Mallophaga, and Corrodentia (as effectively equivalent to Psocoptera) as distinct orders.
Comstock also recognized in time the validity of a separate distinction for the scorpionflies,
recognizing an order Mecoptera (also spelled during the day as Mecaptera), as had Packard and
Hyatt & Arms at about the same time. Indeed, Alpheus Spring Packard (18391905) was extremely
active in the years prior to and during Brauers era. Perhaps his most influential effort was his Guide
to the Study of Insects (108) and its subsequent editions. Therein, as well as in various journal
reports, he set forth a hierarchical, evolutionary classification complete with superorders, orders,
and suborders. He rightly recognized Eurhynchota, which is equivalent to our modern Paraneoptera,
and although he retained them in a superordinal grouping named Phyloptera, Packard made
prescient steps toward the eventual demise of Linnaeuss Neuroptera, recognizing Orthoptera,
Dermaptera (as Dermatoptera), Neuroptera (for Trichoptera and Planipennia), and
Pseudoneuroptera (for Odonata, Ephemeroptera, and Platyptera) (108110), the last with
suborders for the divisions later elevated to orders in their own right. Each of the authors of this last
quarter of the nineteenth century worked to unite diverse character systems, debated different
means of distinguishing homology from analogy, and considered how best to weight conflicting
traits. Insect systematics was becoming more holistic in its sources of evidence, a development that
would be brought to greater fruition in the early twentieth century. The second half of the
nineteenth century saw the rise in influence of paleontological evidence on understanding insect
interrelationships and on the classification of the higher categories. Paleoentomology had been
practiced for many years, but perhaps the most influential fossil seekers were Oswald Heer (1809
1883) and Samuel H. Scudder (18371911), both of whom wrote extensively on the subject, the
latter producing one of the first world catalogs of fossil insects (124). Neither established explicit
phylogenies of the insects, but their more general works were extensively imbued with fossil
evidence and stimulated their contemporaries and followers (45, 129). Interestingly, despite his
particular interest in Paleozoic insects and the obvious challenge of placing them among the

extant taxa, Scudder resisted establishing wholly new orders, preferring instead to establish large,
paraphyletic suborders or other categories, attempting to capture this diversity among the
traditional system of his day. Instead, it was Friedrich Goldenberg (17981881) and Charles
Brongniart (18591899) who first broke that barrier, proposing groups such as the now familiar
Palaeodictyoptera, Megasecoptera, and Protodonata (23), although none contended that there
might have been lineages that did not leave any modern counterparts. Instead, these groups were
regarded as primordial progenitors of our living orders. The said authors influenced their successors
with their perception of the importance and the means of classifying fossil insects. It was Anton
Handlirsch (18651935), director of the Natural History Museum in Vienna, who made the next
major leap in the field, and he was certainly not shy about establishing extinct orders or other
categories of fossil lineages (23, 129). Handlirsch (47) brought the entire subject into focus and made
a critical leap toward fully integrating paleoentomology and neoentomology in his three-volume
monolith Die Fossilen Insekten. Handlirsch not only produced the then most comprehensive account
of the worlds diversity of fossil insects, but also founded a grand scheme of insect evolutionary
history in which paleontological evidence was paramount. Elements of Handlirschs classification
still linger with us today, although some of his ideas were certainly on the fringes, including, for
example, the amazing exhortation that Pterygota arose from a trilobite ancestor (47)

This began a heyday for paleoentomological research, as well as its integration with the study of
living insect diversity. Handlirsch certainly kicked off this research program, but it was the efforts of
Andrey V. Martynov (18791938), Robin J. Tillyard (18811937), and Frank M. Carpenter (1902
1994) that sustained it. All three were avid entomologists in their own right and worked on their
own favorite groups of living insects (interestingly, all three focused mostly on the Mecoptera,
Neuropterida, and Trichoptera). This keen interest in living insects made them adept at interpreting
insect fossils. While Martynov worked on the extensive Paleozoic and Mesozoic deposits of the
Soviet Union, Tillyard developed his own ideas based on the abundance of fossils from Australia and
the United States, particularly the extensive Early Permian deposits of Kansas (45). Indeed, it was
Tillyards series of monographs on the insect fossils from Elmo, Kansas, that helped further galvanize
Carpenters already abiding passion for paleoentomology; in fact, Carpenter would spend the
majority of his 70-year career working on fossils from Kansas and Oklahoma. Both Tillyard and
Martynov proposed explicit classifications of the insects inspired by their paleontological work, even
recognizing additional extinct orders for taxa that did not easily fit within the living lineages.
Whereas some of Martynovs (94) groupings have stood the test of time, those of Tillyard were less
fortunate. Indeed, Tillyards (131) ideas about the phylogeny of insects and the affinities of
particular fossils, such as Protohymenoptera (actually a group of Megasecoptera), fueled
considerable debate even in his own day (21). Unfortunately, both Tillyard and Martynov died in
their late fifties, curtailing their shining careers in entomology and their championing of
paleontological data. While Carpenter did much to highlight the amazing diversity of insects in the
geological past, particularly those from the Paleozoic, he was remarkably reticent to create an
explicit classification based on such lines of evidence. Early in his career, he collaborated on the
second edition of Brues & Melanders Classification of Insects, which was largely a key to families
and less an exhortation of characters supporting a particular arrangement of orders or superordinal
categories (17). Carpenter (22) resurrected Brongniarts discoveries of giant Paleozoic insects with
his discovery of even larger griffenflies (Protodonata) in North America, and highlighted the
importance of characters from such long-lost faunas. Nonetheless, he was a lumper, preferring to
unite the ordinal groups of Handlirsch, Tillyard, Martynov, and other authors rather than use such
evidence to split the fossil lineages into their own categories. Carpenter preferred single,
paraphyletic units from which the still-extant orders derived and used fossil evidence to tweak
individual elements of the classification. Carpenters influence on insect classification was
tremendous and wide ranging, but he articulated only a single, all-encompassing, revised higher
classification in his catalog of fossil insects for the Treatise on Invertebrate Paleontology (23).
Although not working with insect fossils directly, Guy Chester Crampton (18811951) was an eager
advocate of incorporating a paleontological perspective into his intricate morphological studies (28
30), which highlighted, for example, the phylogenetic importance of Grylloblatta, for which he
established the first ordinal name, Notoptera (27). Unfortunately, he was prone to establishing
numerous names for his hypothetical ancestors to particular lineages. Cramptons systems changed
over the years, eventually appearing remarkably modern (Figure 1d ). The influence of paleontology
on insect phylogeny and classification was not uncontroversial. For example, Carl J.B. Borner (1880
1953) did not hide his distrust of Handlirschs heavy re- liance on fragmentary fossils. Borner was
a careful anatomist, and it was he who recognized the Zygentoma (silverfish) as distinct from the
Archaeognatha (bristletails), noting (Figure 1c) that the former were more closely allied to the
Pterygota (12), a grouping Willi Hennig would later dub the Dicondylia. Using similar (mouthpart)
characters, he was the first to consider the Odonata and Neoptera as composing a group, the
Metapterygota (13). For Borner the fossil evidence was simply too incomplete, relying heavily on
wing venation and a meager record (at the time) of bodyemains, as well as too few specimens for
individual lineages, thereby hindering a comprehensive perspective on a particular extinct group.
He did not wholly abandon evidence from the geological past, considering the Dictyoneurida of
Handlirsch to be the progenitor of the Pterygota, but was unwilling to give paleontology its due
beyond this. At about the same time, Richard Heymons (18671943) elucidated the finer aspects of
insect development, discarding the coarse groupings of Ametabola and Paurometabola and instead
providing an arrangement of the orders that emphasized details of their ontogeny (55). His system
split the insects into Epimorpha and Metamorpha, with the former including Epimorpha typica
(apterygotes, Orthoptera, Dermaptera, Isoptera, Psocodea, Heteroptera), Hyperepimorpha
(Homoptera in part), and Hemimetabola (Homoptera in part, Odonata, Plecoptera), and the latter
including Prometabola (Ephemeroptera) and the Holometabola. This early-twentieth century period
in which the aforementioned authors worked was a remarkable time. Amazingly, within a single
decade, four extant higher-rank taxa were described that were immediately or later recognized as
having a pivotal significance in insect phylogeny: Protura (127), Zoraptera (128), Grylloblattodea
(134), and nannochoristid scorpionflies (130). It was also during this same productive period that
the fundamental structure of insect wings was once again considered, and almost simultaneously
three authors recognized the same division and arrangement of orders. Auguste M. Lameere (1864
1942) (85), Crampton (28), and Martynov (93) arrived at the same arrangement, although
Martynovs account is more detailed and the names he accorded these divisions are the ones that
gained almost universal acceptancenamely, Paleoptera and Neoptera. This division, and the
alternative arrangements of the basal winged lineages (Ephemeroptera, Odonatoptera,
Palaeodictyopterida, Neoptera), has remained one of the lasting debates in insect phylogenetics,
one in which Boris N. Schwanwitsch (18891957) first emphasized the arrangement of the flight
musculature and reversed Borners position, with Ephemeroptera as closest to Neoptera (122).
Despite the overwhelming support for Pterygota, Lemche (87) argued for the diphyletic origin of
wings, dividing the insects into the Plagioptera (Palaeodictyopterida and Odonata) and Opisthoptera
(Ephemeroptera and Neoptera) with separate origins. This position was resurrected and expounded
by Matsuda (95) and La Greca (80) but has not garnered further support. Although hypotheses
indicating that the closest relatives of hexapods were members of the myriapod assemblage had
been predominant since the nineteenth century, suggestions of a closer relationship to crustaceans
had been made intermittently, and in 1926, neuroanatomical work led Hanstrom (48) to suggest
that hexapods were subordinate within the Crustacea, thereby foreshadowing much later
developments. Although numerous other contributions and authors debated the relationships of
the orders and their implications for classification, it was not until the middle of the twentieth
century that the next turning point would be reached

WILLI HENNIG, THE CLADISTIC REVOLUTION, AND SUBSEQUENT DEVELOPMENTS

By the mid-twentieth century, the state-of-the-art accounts of higher insect classification were
arguably those presented in Rene G. Jeannels (18791965) (67) chapter on the subject in the
Traite de Zoologie and an article by Crampton (30) summarizing and revising his extensive previous
work. These works are largely congruent with each other (though names applied to supraordinal
groups differ), and the Jeannel chapter, largely reflecting Martynovs views as summarized in a 1938
memoir (94) in Russian, was likely instrumental in making Martynovs conclusions accessible to a
Western readership. It is noteworthy that in none of these works are the nonendopterygote winged
insects grouped together, and indeed Crampton (30) explicitly stated that the usual division of

PHYLOGENETIC SYSTEMATICS AND CLADISTICS The systematizing of organisms by synapomorphy

alone, hence allowing only taxa that are monophyletic in the strict sense (having a history of their
own), was referred to by Hennig as phylogenetic systematics. Because such systematizing is based
exclusively on the branching pattern in the tree of life, it is also called cladistics (from Greek klados,
branch). Some contemporary systematists use the term cladistics in a restricted sense, e.g., for
phylogenies inferred from parsimony-based rather than so-called model-based methods (maximum
likelihood, Bayesian analysis) of character analysis. Arguably, cladistics remains an appropriate term
for all approaches to systematizing organisms according to the relative recency of common ancestry
rather than to some kind of similarity assessment

winged insects into Exopterygota and Endopterygota [concepts introduced by David Sharp (1840
1922) before the turn of the century (126)], on the basis of the external or internal development of
the wings, is utterly meaningless from the standpoint of phylogeny (text in brackets authors own).
In contrast, this division continued to be emphasized in some academia, probably due largely to its
adoption in August D. Immss (18811949) globally influential General Textbook of Entomology
(including its 1957 edition revised by Richards & Davies, 63). More idiosyncratically, Hermann Robert
Weber (18991956) had continued Handlirschs advocacy of endopterygote nonmonophyly, arguing
that beetles, hymenopterans, and the remaining endopterygotes (Webers Neuropteroidea) were
independently derived from nonendopterygote ancestors (135, 136). The core of what has been
called the Hennigian, or cladistic, revolution in biological systematics is the restriction of the
monophyly concept to taxa that include all descendants of their members last common ancestor
and the exclusion of all taxa that are nonmonophyletic in this sense from the formal classification
(see sidebar Phylogenetic Systematics and Cladistics). Although Emil Hans Willi Hennigs (1913
1976) book Grundzuge einer Theorie der phylogenetischen Systematik is often believed to be the
starting point of this revolution (49), his innovative views and terminology were presented more
clearly and more influentially in a major 1953 article (50) on insect systematics, and his ideas
continued to develop during the following years (118, 121). Formal changes to insect classification
were quite limited in Hennigs overall cautious 1953 article, but notable innovations were indeed
made, including discarding such nonmonophyletic taxa as Apterygota and Thysanura (in the sense
of Archaeognatha + Zygentoma) and creating new names for the putative monophyla Dicondylia
(=Zygentoma + Pterygota) and Psocodea (=Psocoptera + Mallophaga + Anoplura). Moreover,
Hennig here tentatively preferred Borners (13) Metaptery- gota (monophyletic Odonata +
Neoptera) solution for the basal splitting event identifiable among extant pterygotes, and he
emphasized the spurious nature of the evidence invoked by Weber (135) for endopterygote
nonmonophyly. The monophyly of the Parametabola (=Paraneoptera), of the Saltatoria
(=Orthoptera), and of Mantodea + Blattodea + Isoptera was accepted, but the monophyly of
Paurometabola (=Polyneoptera) was not, and relationships within this assemblage were left open.
So too were the relationships between Coleoptera, Hymenoptera, Neuropteria, and Mecopteria,
while the Strepsiptera and Siphonaptera remained unassociated with any other endopterygotes. A
monophyletic Eumetabola comprising parametabolans and endopterygotes was tentatively
accepted. The impact of cladistic reasoning on insect phylogeny was modest in the years
immediately following Hennigs 1953 article (52). A contribution on insect phylogeny in Volume 3 of
the Annual Review of Entomology had been commissioned from the entomological polyhistorian
Howard E. Hinton (19131977), but his review (57) was devoted almost entirely to the Mecopterida
(panorpoid orders), on which his empirical work was at that time focused. Other significant
advances made during the past two decades or so to our understanding of the phylogeny of the
Insecta and other classes of the subphylum Antennata were summarized in a series of succinct
statements: (a) the polyphyletic nature of the old groups Myriapoda and Hexapoda; (b) the extreme
isolation of the class Collembola and their lack of any close resemblance to the Insecta; (c) the close
relationship that exists between the Symphyla, Entotrophi (=Diplura), and Insecta ... ; (d ) the Protura
as a class distinct from the Insecta; (e) the very great differences that exist between the
Ephemeroptera and other pterygote insects, including the Odonata; ( f ) the isolation of the
Dictyoptera (roaches, mantises) from the old order Orthoptera and the relation of the Isoptera and
Zoraptera with the Dictyoptera; ( g) the monophyletic origin of the Endopterygota; and (h) the close
relation between the MegalopteraNeuroptera and the Coleoptera-Strepsiptera. Several of these
unreferenced statements were indeed in accordance with then widespread views, but others (a, b,
f: Zoraptera affinities, h) were not. While Hinton in principle endorsed Hennigs methodological
approach, his reiterated advocacy for according micropterigid moths status as a separate order
(Zeugloptera) overlooked Hennigs explicit statements about the proposal [made first by Chapman
in 1917 (25) and subsequently by Hinton in 1946 (56)] being incompatible with phylogenetic-
systematic reasoning. Hintons simultaneous elevation of boreid mecopterans to full ordinal rank
(Neomecoptera) also violated such reasoning, as it was based exclusively on seeming
autapomorphies of Boreidae and hence left out a nonmonophyletic Mecoptera. That some
subsequent findings (138) have supported the boreids closest relatives to be taxa outside the
Mecoptera sensu stricto is another matter. Unsurprisingly, Hennigian cladistics first caught on in
parts of continental Europe where German is the native language or was (then) easily read by most
scholars, and where the approach was communicated to students in introductory university texts
(51, 73). Tuxen (133) applied Hennigian concepts to his 1959 advocacy of the monophyly of
entognathan hexapods, but the center for pioneering empirical contributions to insect classification
was the University of Tubingen, where Weber held the zoology chair from 1951 to his death in
1956. While personally and politically Weber was controversial, it is unquestionable that he and his
students from the 1930s onwards had produced some of the finest-ever descriptive work in insect
anatomy, but he had reluctance, if not downright dislike, of drawing phylogenetic conclusions from
the emerging findings (9, 136). However, inspired by guest lectures at Tubingen from Hennig in the
1960s, one of the last Weber school members, Gerhard Mickoleit, directed his own work, and
subsequently that of numerous research students, toward topics in insect anatomy that appeared
promising for clarifying phylogenetic issues. Early noteworthy examples are Mickoleits (99101)
own studies on the pterothorax, supporting the monophyly of Mecopterida and Diptera +
Mecoptera, and on endopterygote ovipositors and their derivatives, in which for the first time
substantial evidence was provided in support of the overall generalized Neuropterida
(=Raphidioptera + Megaloptera + Neuroptera/Planipennia) as a monophylum, and also for the
sister-group relationship between Neuropterida and the Coleoptera. Moreover, there was
exemplary Weber school work by Achtelig on Neuropterida (1, see additional references in 45), in
which initially putative evidence for Megaloptera paraphyly in terms of Raphidioptera was
presented (a view Achtelig later questioned, 2), and by Rahle (112), who provided unexpected
evidence for a sister-group relationship between the Embiodea and Phasmatodea, a finding later
supported in some molecular analyses. Hennig himself continuously worked on revising/expanding
his account of insect phylogeny, and in 1969 presented the work in book form: Die
Stammesgeschichte der Insekten (53, 54; the posthumously published English translation was
updated and annotated by a group of specialists). Still with many reservations, he proposed a
somewhat more resolved order-level phylogeny.

Sister group: a taxon A that shares with taxon B a common ancestor not inferred to be ancestral to
other known taxa.

in this book than in his 1953 article, including a monophyletic Ellipura, Palaeoptera, and
Paurometabola (=Polyneoptera excluding Plecoptera). A full resolution of the Paurometabola was
also suggested: (Embioptera + ((Notoptera + (Dermaptera + (Mantodea + (Blattariae + Isoptera)))) +
(Ensifera + (Caelifera + Phasmatodea)))). The suggested paraphyly of Orthoptera in terms of
Phasmatodea had been suggested earlier by Sharov (125). The Zoraptera were suggested to be the
sister group to the Paraneoptera (=Psocodea + (Thysanoptera + Hemiptera)), while the
Holometabola/Endopterygota still had much the same modestly resolved arrangement as in 1953.
Hennigs book prompted a review by Kristensen (74), in which a substantial amount of available
literature not considered by Hennig and some original observations were taken into account. The
result was somewhat retrograde in the sense that the evidence bearing on most of Hennigs newly
recognized clades was shown to be more ambiguous/contradictory than evident from his text. A
monophyletic Palaeoptera was considered less probable than a monophyletic Odonata + Neoptera,
monophyly of Orthoptera was tentatively reinstated, but interrelationships within the Plecoptera +
Paurometabola assemblage were otherwise considered unresolved [apart from the long-accepted
monophyly of Mantodea + (Blattodea + Isoptera)]. Kristensen (74, 75) found the placement of the
Strepsiptera to be a particular problem and even found the available evidence for their assignment
to Endopterygota questionable. Soon after, the subject was treated in yet another book. H. Bruce
Boudreauxs (14) Arthropod Phylogeny with Special Reference to Insects provided again a much
(indeed fully!) resolved hexapod tree. Details differed from those in Hennigs proposals: There were
suggested sister-group relationships between Ephemeroptera and Neoptera, between Plecoptera
and Embioptera, and between Grylloblattodea and a clade comprising Zoraptera and (Isoptera +
(Blattodea + Mantodea)). Within Holometabola, the Neuropterida were the sister group to the
Mecopterida rather than to a Coleoptera + Strepsiptera clade. None of the evidence presented for
these changes was considered compelling in subsequent reviews by Kristensen (7577). New
discoveries proved significant in the context of higher classification. The zygentoman Tricholepidion
described in 1961 from California was immediately recognized to have some remarkably primitive
characters (35, 143), but Boudreauxs observation that it retained part of the ligamentous head
endoskeleton otherwise lost in Dicondylia proved particularly challenging. So did the discovery of
the aquatic larva of the small circum-Antarctic scorpionfly genus Nannochorista, in conjunction with
findings on their genital segments in both sexes (102, 103, 111). These taxa were singled out in 1989
(76) as probably representing separate orders (a status actually earlier conferred upon
Nannochorista by Hinton; see sidebar Nannomecoptera); however, although considerable
descriptive and analytical work has subsequently been devoted to them (61, 96, and references
therein), the questions about the sister group of each (i.e., Tricholepidion as related to other
Zygentoma or all other Dicondylia; Nannochorista as related to which other subgroup[s] in the
Mecoptera + Diptera + Siphonaptera assemblage) arguably remain inconclusively answered. in this
book than in his 1953 article, including a monophyletic Ellipura, Palaeoptera, and Paurometabola
(=Polyneoptera excluding Plecoptera). A full resolution of the Paurometabola was also suggested:
(Embioptera + ((Notoptera + (Dermaptera + (Mantodea + (Blattariae + Isoptera)))) + (Ensifera +
(Caelifera + Phasmatodea)))). The suggested paraphyly of Orthoptera in terms of Phasmatodea had
been suggested earlier by Sharov (125). The Zoraptera were suggested to be the sister group to the
Paraneoptera (=Psocodea + (Thysanoptera + Hemiptera)), while the Holometabola/Endopterygota
still had much the same modestly resolved arrangement as in 1953. Hennigs book prompted a
review by Kristensen (74), in which a substantial amount of available literature not considered by
Hennig and some original observations were taken into account. The result was somewhat
retrograde in the sense that the evidence bearing on most of Hennigs newly recognized clades was
shown to be more ambiguous/contradictory than evident from his text. A monophyletic Palaeoptera
was considered less probable than a monophyletic Odonata + Neoptera, monophyly of Orthoptera
was tentatively reinstated, but interrelationships within the Plecoptera + Paurometabola
assemblage were otherwise considered unresolved [apart from the long-accepted monophyly of
Mantodea + (Blattodea + Isoptera)]. Kristensen (74, 75) found the placement of the Strepsiptera to
be a particular problem and even found the available evidence for their assignment to
Endopterygota questionable. Soon after, the subject was treated in yet another book. H. Bruce
Boudreauxs (14) Arthropod Phylogeny with Special Reference to Insects provided again a much
(indeed fully!) resolved hexapod tree. Details differed from those in Hennigs proposals: There were
suggested sister-group relationships between Ephemeroptera and Neoptera, between Plecoptera
and Embioptera, and between Grylloblattodea and a clade comprising Zoraptera and (Isoptera +
(Blattodea + Mantodea)). Within Holometabola, the Neuropterida were the sister group to the
Mecopterida rather than to a Coleoptera + Strepsiptera clade. None of the evidence presented for
these changes was considered compelling in subsequent reviews by Kristensen (7577). New
discoveries proved significant in the context of higher classification. The zygentoman Tricholepidion
described in 1961 from California was immediately recognized to have some remarkably primitive
characters (35, 143), but Boudreauxs observation that it retained part of the ligamentous head
endoskeleton otherwise lost in Dicondylia proved particularly challenging. So did the discovery of
the aquatic larva of the small circum-Antarctic scorpionfly genus Nannochorista, in conjunction with
findings on their genital segments in both sexes (102, 103, 111). These taxa were singled out in 1989
(76) as probably representing separate orders (a status actually earlier conferred upon
Nannochorista by Hinton; see sidebar Nannomecoptera); however, although considerable
descriptive and analytical work has subsequently been devoted to them (61, 96, and references
therein), the questions about the sister group of each (i.e., Tricholepidion as related to other
Zygentoma or all other Dicondylia; Nannochorista as related to which other subgroup[s] in the
Mecoptera + Diptera + Siphonaptera assemblage) arguably remain inconclusively answered.

Clade: shorthand term for groups that are monophyletic in the Hennigian sense

NANNOMECOPTERA Howard Hinton had examined Nannochorista larvae some years before they
were first illustrated (117) and described (111), and in 1967 he was convinced (personal
communication to N.P. Kristensen) that the taxon has nothing to do with Mecoptera, a finding he
intended to publish. However, the formal establishment of the order Nannomecoptera for the genus
only appeared in his posthumously published monograph on insect eggs (58), a most unlikely place
for a taxonomic action of that kind. Unsurprisingly, it was overlooked for many years

Insect palaeontology was actively pursued in the second half of the twentieth century by several
workers, not least in Russia, where work by Aleksandr Grigorevich Sharov (19221973) and Boris
Borisovich Rohdendorf (19041977) was of importance for Hennigs endeavor of establishing times
of origin of recognized high-rank clades (45). Jarmila Kukalova-Pecks (78) extensive
paleoentomological studies led primarily to noteworthy (but controversial; 8) interpretations of
ancestral insect morphology (78). She integrated her paleontological work with observations on
extant taxa to advocate monophyly of Palaeoptera and eventually to establish a neopteran classifi-
cation with a monophyletic Orthoneoptera (=Plecoptera + Embioptera + Orthoptera) and
Blattoneoptera (=Dermaptera + Grylloblattodea + Dictyoptera s.l.) as sister to the Paraneoptera +
Endopterygota (79). An upsurge in palaeontomology was prompted by discoveries of rich new
Lagerstatte (from the Cretaceous in particular) during the last decades of the twentieth century;
unfortunately, Carpenter & Burnhams (24) review of the geological history of insects appeared as
this resurgence was still in its infancy and was understandably lacking. Spectacular new outcrops
provided an abundance of material, while the plethora of insect-bearing amber deposits,
particularly those from the Cretaceous, that accumulated during the 1990s (and continue to accrue
to this day) provided volumes of taxa with enigmatic character combinations, with extralimital
distributions, and with a fidelity of preservation unrivaled in the paleontological world (45).
Paleontological discoveries (e.g., 37, 59, 72), though at times controversially interpreted,
increasingly contributed to the total picture of insect origins and major evolutionary innovations.
When such taxa were cladistically analyzed alongside their modern counterparts, they achieved
their maximum potential for revising concepts of relationships (45). Also, some new concepts of
ordinal monophyly, such as the paraphyly of Blattaria relative to termites (38), found support from
paleontological investigations

EXPANDING CHARACTER SETS

The reexamination of the modest body of evidence drawn upon in attempts to classify high-rank
insect lineages in the 1950s and 1960s highlighted its conflicting nature, and hence the need for
procuring a much more extensive set of comparative data. This was particularly evident in the case
of the assemblage of lower Neoptera (i.e., Martynovs Polyneoptera), which by that time appeared
(and still remains) the most difficult sector in the insect tree of life. Two attempts to systematize
polyneopterous groups using somewhat extensive sets of characters in integumental anatomy and
the then new analytical methods of numerical phenetics were made by Giles (44) and Kamp (68).
Whereas the results of such phenetic analyses had little relevance to phylogenetic systematists,
Blacklith & Blackliths (11) attempt to use a cladistic approach (the early Camin-Sokal method, 20)
was more promising; its informativeness, however, was unfortunately seriously compromised by
the omission of Plecoptera and Embiodea from the analysis. The technical breakthrough in electron
microscopy (EM) (transmission EM in the mid-1950s, scanning EM a decade later) added greatly to
the precision and ease of morphological observation and documentation, but few characters were
thereby added that proved informative for higher-rank insect classification; spermatozoa are the
most notable exception (66). Computer-aided approaches to phylogeny reconstruction, enabling
analyses of sizable character sets, matured during the 1980s, and serious use of molecular
characters in classification of high-rank insect taxa developed toward the end of the decade, initially
drawing almost entirely on ribosomal (r)DNA. Molecular support for an unexpected Strepsiptera +
Diptera (the so-called Halteria) clade led to a much-publicized proposal that the former group
evolved from a dipteran-like ancestor through a homoeotic transformation interchanging the two
pterothoracic segments (140), and the first total evidence study of endopterygote phylogeny (i.e.,
onedrawing on a somewhat sizable morphological data set plus rDNA data) was aimed primarily at
elucidating the Strepsiptera problem (139); here the Halteria again emerged. Although the
Halteria appeared novel, it was actually Lamarck (84) who first classified the Strepsiptera with the
Diptera, highlighting that there is truly nothing new under the sun. The Strepsiptera problem
remained for a while a battlefield between proponents of different approaches (parsimony versus
model-based) to the analysis of molecular data (60), but recent evidence from both morphology and
molecules has brought the Strepsiptera back full circle to being the sister group to Coleoptera (42,
89). A previously unimagined degree of arthropod polyphyly was advocated in Sidnie M. Mantons
(19021979) extensive writings (92) focused on the functional anatomy of feeding and locomotory
structures. With respect to hexapods, a single origin of entognathy was disputed, and the
entognathan orders as well as the Thysanura and Pterygota were claimed to have been derived
independently from a multilegged ancestor whose limbs were nonarticulated (lobopodial), though
it was allowed that the last two could have shared a very early common stage. An earlier
multilegged lobopodial ancestor was suggested to be shared (among extant taxa) only by the
hexapod and myriapod lineages and the Onychophora, and the three were collectively named the
Uniramia. Mantonian views had for some decades a considerable impact on parts of the
international (particularly Anglophone) zoological community, but they were never compatible with
cladistic reasoning, and their influence dwindled as they proved untenable in light of subsequent
developments. Meanwhile, members of the outstanding Dijon school of morphology (the French
Weber school counterpart was founded by J. Robert Denis, 18931969) had procured an
unsurpassed body of factual information, particularly about the primarily apterous hexapods.
Whereas the Dijon workers, like their German colleagues, focused initially on problems in
morphology per se rather than on phylogeny, cladistic analyses of the basal hexapod lineages were
eventually published by Bitsch & Bitsch (10). However, despite the substantial nature of the
analyzed morphological data sets, robust answers to the principal questions (i.e., on the monophyly
of Ellipura, on the closest relations of Diplura, on the position of Tricholepidion) did not emerge.
Only myriapod outgroups were considered in these analyses, but by the 1990s the hypothesis of
Crustacea + Hexapoda monophyly had become revived as a serious competitor to Hexapoda +
Myriapoda monophyly, primarily on the basis of molecular and neuroanatomical data (see
Reference 34 for a summary), and during the following decade it arguably outcompeted the latter.
Developments since the turn of the millennium barely qualify as history (see sidebar Formal Aspects
of Naming). Suffice it to say that the continued and accelerated acquisition of large data sets,
particularly the phylogenomics revolution currently underway, is generating volumes of information
at a scale never before dreamt. Similarly, morphological studies are more imperative than ever (45),
and in some situations, such as with fossils or unique specimens, they can benefit markedly from
novel approaches such as synchrotron and X-ray computerized tomography (36, 81). A recent review
summarizes developments from both molecular and morphological approaches (132). Equally
significant, phylogenetically important taxa continue to be discovered, both as fossils and as living
species. The discovery of living and fossil mantophasmatodeans, likely the closest extant relatives
of the Grylloblattodea, highlights that taxa challenging even our higher-level concepts of
relationships may be found (5, 70). Few of the approximately 10,000 new insect species described
annually (65) will transform our notions of entire orders, but each species does inform us of its
unique combination of characters and biology, refining our knowledge of their close relatives and
thereby generating a cascade of refinements to monophyletic genera, families, superfamilies,
suborders, orders, and beyond. This century will be as dynamic in terms of higher entomological
classification as have been those preceding our current era.

FORMAL ASPECTS OF NAMING The International Committee on Zoological Nomenclature (ICZN)

does not regulate names above the level of the family group (64). Johann Nepomuk Edler von
Laicharting (17541797), however, typified the ordinal names (82). Laicharting established a
classification largely congruent with that of Fabricius but employed names based on the included
genera, nonetheless clearly proposing them as orders (ordo). This resulted in order names such
as Scaraboides (=Coleoptera L., Eleutherata Fabr.), Grylloides (=Ulonata Fabr.), Cimicoides
(=Hemiptera L., Rhyngota Fabr.), Papilionoides (=Lepidoptera L., Glossata Fabr.), Libelluloides
(=Neuroptera L., Synistata Fabr.), Vespoides (=Hymenoptera L., Piezata Fabr.), and Muscoides
(=Diptera L., Antliata Fabr.). A Russian school, which emphasizes the typification of higher names in
the same manner as family-group names under the current ICZN (113, 119), therefore uses as their
starting point Laichartings monograph and even considers these names as coordinate with existing
familial names. Given the essentially universal usage of the ordinal names, particularly those that
have been in common parlance since the work of Linnaeus, the adoption of effectively unknown
typified names serves no good purpose. Arguably, the same applies to the more recently developed
circumscription-based nomenclature (71).

CONCLUSIONS The history of systematics, as in all histories, is not just the story of grand discoveries
or paradigm shifts. It is the sum of all of its actors, toiling to document observations from the natural
world, steadily accruing information that is pieced together into enduring patterns, patterns which
themselves lead us to explanations of greater or lesser degrees of generalization (46). Without the
actions of each and every systematist working passionately on their chosen lineage, the entire effort
would grind to a standstill. No contribution is too small, particularly in entomology where the
numbers of researchers remain meager. However, history also highlights to us the numerous false
starts made as part of the self-correcting body of knowledge we know as Science. These are natural
and we are all a part of this, with todays hubris and certainty becoming tomorrows quaint
anachronism or outright failure. Although some of the assertions and concepts expounded by
authors mentioned herein seem quaint or bizarre (as may some of our own ideas some 100 years
from now!), they were all legitimate hypotheses and served to stimulate the standing intellectual
dialogue. Science needs its variation, even its future failures, just as much as it relies upon the solid
descriptions that underlie, sometimes secretly and unrecognizably, its most powerful theories. As
we hope readers will recognize, the history of entomological classification is a microcosm of such
realities. From the classification we can communicate about many descriptive patterns and
explanatory processes and build predictions, while it itself continues to evolve as new discoveries,
new methods, and new ideas (many certainly doomed to fall away in time) render our applied names
all the more meaningful.

SUMMARY POINTS 1. The history of science is not a steady progression. It has its transformative
moments as well as numerous false starts and reversals. Its entirety shapes the way we think today,
and the present is not an end product but instead merely a chapter in a continuing narrative.
www.annualreviews.org A History of Entomological Classification 601 Annu. Rev. Entomol.
2013.58:585-607. Downloaded from www.annualreviews.org by NORTH CAROLINA STATE
UNIVERSITY on 01/15/13. For personal use only. EN58CH29-Engel ARI 28 November 2012 20:12 2.
Pre-Darwinian authors were astute and keen observers of the natural world, recognizing enduring
patterns of characters and, more often than not, correctly interpreting and classifying the species
before them. All this work was done with quaint tools and with the tiniest fraction of material
compared with what we have today. 3. The Darwinian paradigm shift transformed how authors
interpreted their observed and described patterns of characters, but it did not substantially
augment how they actually practiced systematics. Classifications were changing wildly, but this was
not due to a tectonic shift in the day-to-day practice of the science. Evolution was a conceptual and
explanatory change long before it became methodological. 4. With the rise of cladistic reasoning
and numerical practices, systematics experienced its greatest methodological shift. Emphasis on
monophyly and more precise estimates of genealogical relationship gave classifications far more
predictive power than they ever had before. Simultaneously, the diversity and magnitude of data
sets have grown exponentially, enlivening the field and strengthening the overarching classification
of insects.

DISCLOSURE STATEMENT

The authors are not aware of any affiliations, memberships, funding, or financial holdings that might
be perceived as affecting the objectivity of this review.

ACKNOWLEDGMENTS

We apologize to those entomological systematists whose efforts over the centuries have inspired
and guided us all, but for whom we were unable to devote space to the memory of their
accomplishments. We are grateful to I.A. Hinojosa-Daz (University of Kansas) for assistance with
images and to Ted R. Schultz for inviting us to provide this overview, stimulating its production and
providing constructive criticism.

14. Next-generation successor to Hennigs 1969 book, with many independent views on evidence
bearing on interrelations of high-rank taxa.

16. Substantial pioneering application of evolutionary thought to hexapod classification

30. A summary of a lifetimes work on hexapod morphology and evolution, representing a
precladistic state-of-the-art insect classification (see also 94).

45. A modern standard reference on insect evolution and classification, integrating paleontological
and neontological evidence from a cladistic perspective.

47. A monolithic synthesis of long-lasting impact integrating the paleontological and neontological
evidence available at its time

50. Benchmark application of Hennigs principles of phylogenetic systematics to hexapod

classification.

53. Expanded Hennigs previous writings on hexapod phylogeny, focusing on the time-of-origin of
putative monophyla.

74. Simultaneously brought Hennigs (53) classification attention to a wider audience and provided
greater clarification of the critical issues in insect phylogeny.

94. Synthetic account of paleontological and neontological evidence, representing a precladistic

state-of-the-art insect classification (see also 30).

137. The entomological text of its day, paving the way for insect classification in the Darwinian era.