Several versions of ‘social Darwinism’ flourished in the late nineteenth and early twentieth century, along with ideologies derived from non-Darwinian evolution theories. They exploited discoveries of fossil hominids including Neanderthals and the Piltdown fraud to construct rival explanations of the emergence of human characteristics that might shape social development. The linear hierarchy of races erected in the nineteenth century remained the basis of many popular accounts, even though professional anthropologists began to turn their backs on it. Ideologies based on national or racial competition were advocated even by writers who did not accept the Darwinian theory of competition within populations. Fear of racial degeneration fuelled the eugenics movement’s calls for the elimination of ‘harmful’ characters, although the input from genetics encouraged an analogy with artificial rather than natural selection.

How did prehistoric humans live during the Paleolithic?

Humans are the end point of almost four billion years of evolution. The generally accepted explanatory theory is Charles Darwin’s theory of evolution through natural selection, brought on by a struggle for existence. Darwin bolstered his case through a “consilience of inductions,” showing how evolution through selection explained in many areas right across the life sciences – behavior, fossil record, geographical distributions, morphology, taxonomy, and embryology. The missing element was an adequate theory of heredity, supplied by the Moravian monk Gregor Mendel. A major problem was why random change in the units of heredity, mutations of the genes, should provide enough suitable material for selection to work on. Theodosius Dobzhansky solved this problem by hypothesizing that all populations contain much variation, quite enough for selection to work on. His student, Richard Lewontin, using molecular techniques – gel electrophoresis – showed that Dobzhansky’s hypothesis was right.

Chapter 3 unpacks the jewellery inference about Neanderthal language and appraises its soundness. Including three inferential steps represented by arrows, this inference looks in outline as follows: The Neanderthal occupants of the archaeological sites S1,…,Sn were associated with the objects O1,…,On → They wore these objects as personal ornaments → They treated these ornaments as symbols → They had language. [Sites S1,…,Sn include the Grotte du Renne and at least seven other caves/shelters. The objects O1,…, On include, inter alia, marine shells, raptor talons and feathers and perforated animal teeth.] With reference to a large literature, Chapter 3 finds the following about the soundness of the jewellery inference: (a) the data from which the first inference starts provide adequate empirical grounding for it; (b) this step is warranted by an accepted theory of the distinctive propeties of personal ornaments; (c)the second step is unwarranted: it is not underpinned by an adequate theory of the distinctive properties of symbols as opposed to other signs. This finding implies that the third inferential step lacks the necessary grounding.

Chapter 6 elucidates and appraises the deliberate-burial inference about Neanderthal language. In outline, this inference comprises the following three inferential steps: Data about the Neandertal skeletal remains R1,…,Rn found in the state C at the caves C1,…,Cn along with the objects O1,…,On → These Neanderthals were deliberately buried → The Neanderthals who carried out the burials behaved symbolically → These Neanderthals had language. The first inferential step is found to be well warranted by an accepted theory of the properties of deliberate Neanderthal burials. One of these properties is that the skeletal remains are located in a natural or Neanderthal-made pit. For the second inferential step to be empirically grounded, the skeletal remains are required to be accompanied by grave goods, objects ritually deposited along with the body for use in after life. There isn't evidence that the objects – e.g., flint scrapers, the upper jawbone of a red dear, a rhinoceros tooth – found along with Neanderthal skeletal remains were indeed grave goods. The second inferential step is, accordingly, considered unsound in the literature, leaving the third ungrounded.

Chapter 10 appraises the hunting inference about Neanderthal language. Starting from data about Neanderthals’ ambush hunting, this inference comprises in outline three inferential steps: Neanderthals’ ambush hunting had the features F1, F2, F3 and F4 → This hunting required cooperation → This cooperation required communication → This communication required language. [F1 = The hunting was confrontational; F2 = The prey comprised large animals; F3 = Heavy carcasses had to be transported; F4 = Bands of hunters were divided into a tracking/driving and an ambushing group.] The first inferential step is underpinned by the assumption that hunters had to cooperate in planning ambushes, killing dangerous prey etc.; the second step by the assumption that such planning required communication about effective hunting strategies, coordinated action etc.; the third step by the assumption that such communication required ‘labels’ or ‘words’ for referring to features of the landscape, routes, hunting sites etc. These assumptions are reasonable. But it is unwarranted to infer that (a) to be successful, such communication required complex grammar; and (b) that any such communication was not gestural.

Chapter 5 analyses the body-decoration inference about Neanderthal language. This inference includes the following three inferential steps, depicted by the arrows: Blocks of manganese dioxide yielding black pigment were associated with Neanderthal ocupants of sites S1,…,Sn → This pigment was used by these Neanderthals to decorate their bodies → These body decorations had a symbolic function for these Neanderthals → These Neanderthals had language. [Sites S1,…,Sn include Pech d l'Aze I and Pech de l'Aze IV in southern France.] Chapter 5 goes into doubts about the first inferential step of this inference. The main one is that it has not been excluded that the Neanderthals concerned used black pigment for an utilitarian instead of a symbolic function. Black pigment is known to have been used ethnographically and prehistorically for many utilitarian functions. And, importantly, there is experimental evidence that the Neanderthals at Pech de l'Aze I used manganese dioxide for fire-making. Until seccessfully challenged, this finding represents a good reason for doubting the soundness of the second inferential step. This leaves the second and third inferential steps ungrounded.

Chapter 4 elucidates various inferences about Neanderthal language drawn from so-called cave art attributed to Neanderthals. Skeletally, these inferences look as follows, arrows depicting inferential steps: The markings M1,…,Mn are found on the walls of the Iberian caves C1,…,Cn → These markings represent art made by Neanderthal occupants of the caves → This art had a symbolic function for these Neanderthals → These Neanderthals had language. The markings include a hashtag engraving, red disks and hand stencils, a red ladder-shaped sign and red painted mineral deposits. Serious concerns have been expressed about the soundness of these inferences. Two are fundamental. First, the empirical grounding of some are suspect: the dating of the markings is claimed to be inaccurate. This means that some markings may have been made by a modern human rather than a Neanderthal. Second, it has been pointed out, the meanings of the markings are a mystery. This implies that it is unwarranted to infer that these markings were symbols. They could have had a non-symbolic function, which would make the third inferential step ungrounded. Chapter 4 discusses these and other doubts at length.

Chapter 2 sets out the conceptual tools used in the book for analysing selected inferences about Neanderthal language. Derived from The Windows Approach to language evolution, these tools include conditions on the soundness of inferences drawn about something from data about another thing. Such inferences are not necessarily sound. The chapter illustrates three fundamental soundness conditions, using them in an appraisal of a composite inference about the linguistic capacity of European Neanderthals drawn from data about scratches in a number of their anterior teeth. These conditions are: (a) An inferential step must be grounded in factual data; (b) An inferential step must be underpinned by a warrant; and (c) A conclusion must be pertinent, referring to clearly identified and correctly characterised entities. Depicted by arrows, the four steps of the scratched-teeth inference are the following: There are scratches in a number of anterior Neanderthal teeth → Neanderthals ate with the right hand → They were right-handed → They had left-lateralised brains → They had linguistic capacity. The chapter finds the third inferential step to be contentious, and the fourth to be unsound.

Chapter 9 focuses on an inference about Neanderthal language drawn from data about the experimental teaching of stone-tool making to modern humans. Comprising three inferential steps, the knapping-pedagogy inference looks as follows in outline: Experimentally gathered data about the transmission of Oldowan technology to modern humans → Verbal language enhances the transmission of Oldowan technology to modern humans → Verbal language originated hundreds of thousands years ago as a prerequisite for Acheulean technology → Some Neanderthals had verbal language. There are various reasons for doubting the soundness of this inference. First, its empirical grounding is contentious: some experimental studies have found verbal interaction to be unnecessary for teaching knapping skills, even causing underperformance by modern learners. Second, Chapter 9 finds the inferential steps to lack the required logical force. For instance, it cannot be inferred that verbal language was a prerequisite for teaching knapping technology to modern humans from data that it only enhances the teaching of such technology to modern humans. This logical flaw leaves the final inferential step ungrounded.

Chapter 11 elaborates on a number of points argued in preceding chapters. A first concerns the finding that the inferences drawn about Neandertal language from the Neanderthal behaviours at issue are less than sound. This does not imply that Neanderthals could not have had a form of language. Nor does it imply that non-behavioural attributes of Neanderthals provide better windows on their linguistic attributes. This is substantiated by an appraisal of two important inferences about Neanderthal language drawn from putative correspondences between the brains and genes of Neanderthals and those of modern humans. Both the gene and the brain inference are found to be unsound. A second point concerns the roots of the controversial nature of much that has been claimed about Neanderthal language. An often cited one is the lack of uncontentious evidence about Neanderthal language. The root cause, however, lies deeper in a poor conceptual framework that lacks, inter alia, appropriate conditions on the soundness of inferences. A third point concerns the mysterious nature of Neanderthals’ language. Chapter 11 argues that credible inferences about what it involved can be drawn from Neanderthals’ cooperative hunting. It comprised referential signs but lacked complex grammar.

Chapter 1 sets out the concerns, aim, approach and focus of the book, and lists some of its major findings. Its guiding concern is with the lack of uncontroversial answers to questions about the existence and nature of Neanderthal language. Its maim aim is to disperse the murk veiling this phenomenon. For pursuing this aim, the book adopts The Windows Approach to language evolution. In terms of this approach, for claims about Neanderthal language to be credible, they must be conclusions of sound inferences about it. The book, accordingly, unpacks a range of important inferences about Neanderthal language and appraises their soundness. These inferences are drawn from four allegdly symbolic and three non-symbolic behaviours. The book restricts its focus further to inferences about language as a cognitive capacity as distinct from speech as a form of behaviour that presupposes language. To guage the heuristic potential of Neanderthal behaviours as windows on Neanderthal language, the book apparaises in addition the soundeness of important inferences drawn about Neanderthal language from two biological attributes of Neanderthals: the FOXP2 gene and the Broca’s area of the brain.