1. Introduction

For most of the last quarter million years, Europe and a good part of Western Asia have been home to an enigmatic and curious race called Neandertal. Neandertals were first discovered in the middle of the 19th century: the Engis child from Belgium (1829), Forbes Quarry from Gibraltar (1848, but not recognized as particularly special until 1865), and of course the Neandertal skeleton from the Feldhofer cave (1856) in the Neander valley near Dusseldorf. Neandertals are the only fossil human group whose name is commonly used in modern language; it is no compliment to be called a Neandertal.

The "Neandertal question" - the place of the Neandertals in human evolution - is paleoanthropology's oldest problem, and its solution was astonishing. It may be the most famous example of a controversy of evolution or extinction from the fossil record.

2. The first interpretations

Neandertals were a problem for natural philosophers of the 19th century because they thrust the possibility of a savage and primitive European ancestry right in the face of the European scientific world – these Neandertals were the "noble savages," but the wrong ones, of Thomas Hobbes and Charles Dickens and far more savage than noble. This was a time that combined European exceptionalism and eurocentralism, when savagery and primitiveness were not expected for Europeans because they were found someplace else, in places far from Europe that the Europeans had conquered.

The prevailing phylogeny, actually the first phylogeny, was that of Ernst Haeckel, Darwin's Prussian disciple. In 1868 Haeckel used a tree to represent the phylogeny of species as a branching picture of their evolutionary history. The phylogeny he created in the 1860’s to show the place of humanity in the natural world postulated a hypothetical ancestor between ape and man that was called "Pithecanthropus alalus", speechless ape-man, this well before Eugène Dubois discovered the fossils he named Pithecanthropus, from Trinil in 1891. The fate of how Neandertals were envisioned was sealed at this time; before their fossils could be actually placed on Haeckel’s human family tree between Pithecanthropus and Homo sapiens, Haeckel had labeled the position for them as "Homo stupidus."

Hermann Schaaffhausen gave the first evolutionary interpretation of a Neandertal, the Feldhofer specimen (1859). He considered the Neanderthal as an archaic (‘lower’, in his words) form of human being. The nineteenth-century argument developed about whether the Feldhofer Neanderthal skeleton was pathological or an ancient and primitive race (muddying the issue, we now understand that Feldhofer is both pathological and archaic). William King gave the Neandertal a unique species name in 1864; after that the Neandertal problem wouldn’t go away, because what began as just a few Neandertal fossils grew to become a sample of so many that it quickly was the largest sample of fossil humans known from anywhere, and Neandertals predominated ideas about human evolution for decades.

With the Neandertals’ practice of living in south-facing European caves and burying their dead in them, the sample grew quickly once paleoanthropologists realized where and how to look for the Neandertal remains. Ironically, this resulted in a much larger sample size of Neandertal fossils, relative to other fossil human remains, than the actual very small size of the Neandertal population relative to the size of the rest of penecontemporary humanity. As it turned out, the number of fossils was related to many things, first and foremost the opportunity for burial soon after death, but had nothing to do with the number of Neandertals who once lived. Despite a large number of Neandertal fossils, we now understand that the number of Neandertals actually alive at any time was very, very small. Indeed, the total size of the human species probably approximated a million for most of the Pleistocene (Eller, Hawks & Relethford, 2004).

3. No Neandertal ancestor

The easy solution to the conundrum of recognizing Neandertals as primitive Europeans was that Neandertals must not have been European ancestors; in this case, it would not matter that they were primitives. While a minority of 19th and earlier 20th scientists, especially from the German-speaking world of Central and Eastern Europe (Radovčić 1988), thought of Neandertals as a part of humanity, with a few exceptions most western European and American scientists accepted the premise proposed by William King when he gave the Feldhofer Neandertal a species name because "so closely does the fossil cranium resemble that of the chimpanzee, as to lead one to doubt the propriety of generically placing it with man." He named the Neandertal species "Homo neanderthalensis¹ King," showing Neandertals as a separate branch on the human family tree, and not as a European ancestor. A separate branch for Neandertals meant they must have become extinct.

A Neandertal species made sense at that time. The bones from the Feldhofer Cave were clearly unusual, and the 1908 discovery of a complete Neandertal burial at Chapelle-aux-Saints in France allowed Marcellin Boule (1913) to systematically describe the "Old Man" (Fig. 1) and compare him with modern humans and chimpanzees, the closest primate relative of humanity (Fig. 2). For the most part Boule showed that the Neandertal was midway between chimpanzees and humans, but had a few unique features that were unlike either of these living species. Boule presented this detailed and systematic report as strong evidence that Neandertals were a distinctive line of human evolution closely related to humanity, but a different species, as King had suggested from far more meager evidence, which ultimately became extinct. The virtual deluge of Neandertal discoveries that soon followed (Trinkaus & Shipman, 1993) were mostly interpreted in this framework, and today it is common to find scientific literature discussing the causation, dating, and evolutionary details of when and how the Neandertal species came to be, persisted, and disappeared (e.g. Stringer & Gamble, 1993 ; Tattersall, 1995). The Piltdown fraud played a key role in this interpretation, until the mid-century when it was shown to be a hoax (Radovčić, 1988). As long as Piltdown was accepted as a valid, accurately dated human ancestor, Neandertals could not possibly be an ancestor.

Even after the Piltdown fraud was exposed (Weiner, Oakley & LeGros Clark, 1953; Russel, 2003), Boule’s thinking remained influential and many paleoanthropologists from mid-century on sought differences in the Neandertal fossils that demonstrated they were a different, extinct species. Intellectual successors included prominent influential paleoanthropologists from many countries such as Americans Ian Tattersall and F. Clark Howell, Chris Stringer (British), and French scholars such as Henri Vallois and Jean-Jacques Hublin.

This was certainly the popular opinion, seen in cinema and theater as well as in popular books, magazines and the press whenever there was an excuse to discuss Neandertal issues. For instance, in Everyday Life in the Old Stone Age, a popular book from the middle of the last century, Marjorie and Charles Quennell wrote: His large head, with the thick frontal bones, must have been very good for butting a brother Neanderthal, but it was no use against the stone wall of advancing civilization, and like the Tasmanian and Bushman, the Red Indian and Australian of nowadays, he fades out of the picture and his place is taken by a cleverer people. Neandertal extinction had become a metaphor for all colonialist invasions and occupations.

4. Central Europeans Differed

There always was a group of scientists who saw Neandertals differently, as part of the human species. In the 20th Century German-speaking world, these included Franz Weidenreich (1943) from the German Palatinate, and two Czechs: Aleš Hrdlička (1929) and Jan Jelínek (1969). Hrdlička and ultimately Weidenreich immigrated to the United States, where Hrdlička founded the American Association of Physical Anthropologists and its journal. It is an unfortunate quirk of fate that each of the three worked in institutions or museums where they had no direct intellectual influence as the teachers of generations of students.

5. The Evolutionary Synthesis

And then the tide began to turn.

Views of human evolution came to incorporate the evolutionary synthesis, developed in the late 1930’s and 1940's by a group of evolutionists, paleontologists, and geneticists seeking to reconcile views of evolution as slow adaptive change, with views of evolution as driven by random mutations. One great change in perspective that came with the evolutionary synthesis was a shift for all biological sciences, including biological anthropology, from thinking of the individual to thinking of the population as the basis for the evolutionary process (Washburn, 1951). Considering Neandertals as a population was a change in framework that promoted different kinds of questions about their biology and behavior, as well as their origins and evolutionary fate.

In 1939 a major comprehensive monograph was published on the ancient human burials from the Mount Carmel Caves in the Levant of Western Asia. This was the first sample to be examined. In it, the authors, Theodore McCown and Arthur Keith, recognized that some of the specimens could be considered Neandertals and others moderns (or Cro-Magnons), but that the terminology² was a poor representation of the variation because some modern traits were found in the Neandertals, and some Neandertal traits were found in the moderns (Fig. 3). They took the position that, at least locally, the Neandertals were the ancestors of the moderns and interpreted the Mount Carmel sample as being "in the throes of evolutionary change," explaining its exceptional variation as a snapshot taken as Neandertals evolved into moderns. Theodosius Dobzhansky, a Russian émigré geneticist and one of the architects of the evolutionary synthesis, applied newly understood evolutionary concepts to the Mount Carmel question in 1944, re-analyzing the Mount Carmel variation and interpreting it differently, as the result of populations of Neandertals and moderns mixing. The fact of mixture showed that these two populations were penecontemporaries in the same species, and Dobzhansky thereby identified Neandertals as a human race or subspecies (these have the same meaning), a view that until then was only commonly held by Central and Eastern European scientists. For Dobzhansky, and many later evolutionists (for instance Ernst Mayr, in his 1950 reevaluation of human taxonomy) this evidence of mixing showed that Neandertals could be recognized as part of H. sapiens.

6. Human Races Do Not Exist

Today, it is almost universally agreed that human beings do not have a pattern of biological variation that reflects subspecies, or races (Caspari, 2010; Marks, 1995; Wolpoff & Caspari, 1997). Human races exist in a social sense and have powerful, often disturbing, influences in how societies are organized and how individuals behave. This sense of race as a social construct is real, but its reality does not reflect biological difference. Human biology does vary geographically, but there are not groups of people with distinct boundaries and unique populational histories – genealogical aspects of variation we would expect for geographic variation to be racial variation. Conversely, as Rachel Caspari notes, "while social races are not genealogical entities, they have biological dimensions."

The fact that there are no living human races does not mean there were never human races in the past.

7. Recent Neandertals Show Mixture

In the Levant, better dates and more accurate archaeological assessments show the Neandertals and other humans were present at the same time, for a much longer time that the Mount Carmel caves, and there was ample opportunity for mixing, Dobzhansky’s interpretation was valid. But this is not the case in Europe, where every identifiable Neandertal is earlier than every identifiable "modern". Here the issue is about whether or not transitional specimens were evident at late Neandertal sites such as Spy in Belgium (Semal et al. 2009), and Fontéchevade (Chase et al. 2009) and St. Césaire (Vandermeersch, 1984) in France, all approximately 36,000 years old. Each of these shows evidence of mixture. St. Césaire lacks many "classic" Neandertal features although it is clearly related to Neandertals. In fact what St. Césaire most closely resembles in the parts preserved is the "early modern" (neanthropic) Levantine, Skhul 5 (Fig. 4). At Spy two adult crania and skeletons were discovered in1886, these complete burials were poorly excavated and the skeletons mixed up in transport to the laboratory.

Even still, the specimens revealed that there was significant variation in cranial shape and the supraorbital torus. Spy 1 is very similar to other Neandertals but Spy 2 is more rounded, had a more bulging forehead, and in other ways is quite like some of the early modern Europeans (Fig. 5). At Fontéchevade there is a long-standing controversy about the shape and development of the supraorbital torus in the missing part of the frontal, in an otherwise very Neandertal-like vault.

Specimens such as these could conceivably be cases of late Neandertals evolving into a population more like modern Europeans, or of Neandertal mixing with populations entering Europe as the climate improved during the period between glacial advances. Either of these explanations requires that Neandertals are members of Homo sapiens who contributed to modern human variation, and of course they could both be true. Further examining this question, I worked with several colleagues (Wolpoff et al. 2001) to test the hypothesis of dual ancestry for early modern central Europeans from the Mladeč cave in Moravia, a contention of multiregional evolution.. The complete replacement theory (Eve theory), then widely believed, meant that early modern Europeans could only have one ancestor, Levantine populations, some of African descent, entering Europe during the interstadial period between glacial advances. However, we concluded that Mladeč had more than one ancestor because we could not disprove a hypothesis of its equal relationship to both of these older populations, Levantines and European Neandertals. This would account for the anatomy of the transitional late Neandertals from Spy, St. Césaire, and Fontéchevade; they were mixing with other populations entering Europe before the warmest part of the interstadial, and not necessarily all from Africa (Otte).

8. Ancient Nuclear DNA

The final nail in the coffin for Neandertal extinction came with reading the nuclear DNA (nDNA) in Neandertal fossils. Reading the nDNA of several Neandertals and reconstructing a full Neandertal nuclear genome, the most startling discovery in 21st century paleoanthropology, provided direct evidence of alleles in the gene pool today that are Neandertal-derived (Green et al. 2010). Many of these have been under recent selection.

There are two consequences of this new work that address the Neandertal issue. The first is that Neandertals interbreed with people entering Europe from Africa. Neandertal-derived genes make up, on average at least 4-5% of the genome in Europeans and Asians, but are not found in Africa. It is thought that much of this interbreeding was in the Levant, where emerging Africans are known to have encountered Neandertals. These results verify anatomical studies showing that later Europeans have some Neandertal ancestry (Fig. 6). But Europeans are not the only population to inherit Neandertal genes. For the most part the Neandertal genes in Asian populations are different from the Neandertal genes in European populations. This reflects the differing histories of drift in what must have been very small populations.

Secondly we learn much about ancient population structure from instances of introgression (Hawks & Cochran, 2006), cases of an ancient allele from an archaic population entering population ancestral to living people much more recently than the time of the mutation giving the allele. Introgression demonstrates a time of mixture with the ancient population. The introgressing alleles must have been under selection in populations where they first appeared, because they quickly increased in frequency when they entered populations ancestral to living people, as the gene flow restrictions of the Pleistocene reduced, and finally collapsed.

Then, nDNA from a non-Neandertal specimen from Denisova Cave in Siberia (Reich et al. 2010) provided further evidence of gene flow between archaic hominids and other ancestors of modern humans. This is a different lineage than the Neandertal one, the two share a common ancestor and probably exchanged some genes, but descendent genes unique to the Denisovans have only been found in Melanesians so far. Combined with Neandertals these ancient populations are estimated to account for some 8% of the ancestry of living humans (Reich et al. 2010). This is a substantial percentage, given that these peripheral populations (Neandertals and Denisovans) at the northern margin of the Eurasian range were quite small (Eller, Hawks & Relethford, 2004, as noted above) and may not have been as much as 8% of the world population when they lived. Neandertals and Denisovans must have had significant restrictions in gene flow from the Africans, as well as from each other, until the time of recent mixture. This is direct evidence that while there was restricted gene flow, during a time of very small size for the human species when the somewhat separated populations were subspecies (or races) as Dobzhansky had thought. Introgression does not happen today, there is a very much larger human population and no gene flow restrictions to prevent favorable alleles from dispersing everywhere as soon as they appear. What changed the ancient human population structure to the modern one was the very large population size increases beginning about 50,000 (Hawks et al. 2007) and really evident in the last 10,000 years ago (Gignoux, Henn & Mountain, 2011).

Recognizing the possibility that Neandertals and Denisovans were past human subspecies, or races, almost certainty two of a larger number, opens up new ways to think about the biological relationships of Neandertals to other past populations, and to ourselves, the survivors.

9. Summary

Evolutionary theory, fossil evidence, and nuclear DNA sequencing have finally combined to show that Neandertals did not become extinct without descendents (Fig. 7).

As a distinct population, Neandertals are gone, but many of their genes remain and these have significant influence on the variation in some living polpulations. Neandertals are one of three known human races, or subspecies, from the Pleistocene – there almost certainly were more. The population structure of the human species was different then from what it is now (Wolpoff, 2009), for the most part because of very small population size, but then as now all human varieties were equally human; they exchanged genes, so that really important anatomies and behaviors such as language abilities, and brain complexity and structure as reflected in brain size (Lee & Wolpoff, 2003), could disperse under selection throughout the whole species (both Neandertals and Denisovans share a unique human gene related to language production that is not as old as their estimated common ancestor). Gene flow was restricted (Templeton, 2007), not absent. If there is any generality that cam be learned from this brief history, perhaps more than anything else it shows that for the human species, the present is a very poor guide for understanding the past.

Notes

¹ "Neanderthal" is the older German spelling of the valley where the Feldhofer cave is located. A change in German spelling at the turn of the last century dropped the "h" and the place is now called "Neandertal". The older spelling, Neanderthal, really should only be used for the Feldhofer fossil, or the sample of Neandertal fossils, by those who regard these folk as a distinct species (Homo neanderthalensis). Species names are carved in stone and cannot change even when common spelling does, and so using the older spelling signals that the Neandertals referred to are Homo neanderthalensis.

² McCown and Keith substituted a different terminology, Palaeoanthropic and Neanthropic, but this did not really solve the problem.