If I can go back to word order, I can offer a very different view...

As Gell-Manna and Ruhlen (2011) note, the reconstruction of an SOV stage “need not necessarily refer all of the way back to the time when behaviorally modern humans emerged and peopled the Old World. There could have been a “bottleneck” effect at a much later time, with a single language spoken then being ancestral to all or most attested languages.”

If my syntactic reconstruction is on the right track, then the initial stages of syntax were intransitive and absolutive-like, without a subject/verb distinction, and as such could not have had either SOV or SVO structures, or any other type of transitive structure. Instead, this would have been a two-word stage with e.g. a verb and one noun (VN or NV) or perhaps, if this single argument is taken to be a subject regardless of its thematic/semantic role, then we can say the order was either VS or SV. But no objects. Sort of what happens with intransitive absolutive structures in absolutive languages, but also in certain constructions in nom-acc languages (e.g. middles; certain compounds, etc.). One example that can illustrate this stage would be VerbNoun compounds like: scare-crow, kill-joy, pick-pocket (where the noun seems object-like: one who kills joy), and rattle-snake, cry-baby, busy-body, stink-bug (where the noun seems subject-like: snake that rattles), and then e.g. dare-devil (where it seems it can go either way: one who dares devil or devil who dares; native speakers in fact disagree re which one of the two is right). Notice that the syntax/structure of these compounds is rigid and very constrained (two-slot grammar, with the verb first, and the noun second, regardless of its role to the verb). If language started in a comparable fashion, from the very beginning, there would have been some autonomy of syntax, that is, the syntax would constrain what and how much we can express in a single utterance. Even though semantically we can perceive an event as involving an agent, a patient/theme, an instrument, a location, a benefactor, if the syntax mold had only two slots, then one could express only one of these roles, leaving the rest up to pragmatics. Just like with the compounds above. Note that Givón (1979) proposed that there was a presyntactic stage which had a “low noun to verb ratio.” In my analysis, this was a syntactic stage, although of the simplest possible kind.

In this view, this is not to say that semantics and pragmatics did not play a role, and that Agent First was not relevant, but it is to say that these principles did not determine the outcome of syntax, or the ordering. Rather, from the very beginning, continuing to this day, there has been a tension between syntax (and the number of slots it provides in any given construction), and the semantic/pragmatic possibilities. That seems more consistent with the variety of word order “solutions” across languages, and within single languages as well.

Furthermore, if the initial stage of language was this absolutive-like (VS or SV) syntax, then transitivity could have arisen via parallel paratactic concatenations of the kind: Woman push, man fall, as have been reported for the early stages of the emerging sign languages, and as also attested with serial verb constructions available in various languages, either in that order, or in the Woman push-fall man order. What they have in common is one argument per verb (intransitivity). If this is true, then Agent-First as discussed by many can in fact reduce to Cause-First, which is a more comprehensive principle, and which operates across languages in paratactic constructions such as: Nothing ventured, nothing gained. No justice, no peace. No pain, no gain. Easy come, easy go. These do not involve agents or transitivity, but the cause has to precede, as we cannot reverse the order. If so, then Agent First in e.g. Woman push man fall can be a consequence of the more general Cause First strategy.

In any event, I am trying to say that my reconstruction of syntax is not consistent with the human proto-language starting with some kind of natural-seeming transitive word order, whether SOV or SVO. I offer many reasons for that in my book, including an astonishing variation across languages, and within languages, regarding the expression of transitivity. I have attached one chapter of the book that outlines the basic arguments.

An old book that may serve as a good starting point here is the classic, The History of Language, dear Pablo.

Hello, Marwan,

Thank you for the hint. I am familiar with the classics, though. It's 'non-classics' that I miss, as they might help to unblock the deadlock in the language evolution discussion, by bridging the gap between the nativist and the emergentist strands. Data overflows, views abound, but the dichotomy persists—that is, so long as we still take account of the views of Hauser, Fitch, Chomsky, Jackendoff, Pinker..., which some no longer do. Or perhaps the divergences within one and the other strand are too large by now... So, rephrasing my question, I would add in the first place: do you find such a dichotomy persisting thus far? 

Regards

Hi Pablo, IMO these theoretical discussions (esp.Chomsky) have largely become irrelevant. Dobshansky: "Nothing in biology makes sense except in the light of evolution." If we analyze human language into more elementary parts, and compare these to other animals, speech & language origins become much less mysterious IMO.

Music, rhythm, variable tones & loud sounds were already present in gibbon duetting (LCA c 18 Ma?). Much later (probably early-Pleistocene) Homo spread intercontinentally along the coasts between Algeria, the Rift, Georgia & Java (Aïn-Hanech, Turkana, Dmanisi & Mojokerto resp., all 1.8 Ma, and coastal or connected to the coasts), where their diet included littoral foods. The combination of gibbon-like song (early hominoids) + adaptations for seafood collection (pre-sapiens Homo) provided the necessary elements for speech evolution:

- Voluntary breathing control arose when Pleistocene Homo (

Hi Marc,

So is it in my opinion, too. Cognitive, anatomical and other biological factors indeed seem the best-grounded by far, though the roles of social interaction and cultural 'ratcheting' are not to be forgotten either (cf. the 'biocultural hybrid'; see Tomasello, 2008; 2009; and Levinson & Holler, 2014). 

In preparing an overview of this inquiry lately, I have gone over dozens of accounts; however, I haven't come across any of those takes on seafood in evolution. This makes sense: one conclusion of my overview is that these alternative, (more) profound analyses often go unnoticed, below the mainstream stands on the abovementioned dichotomy. This bias is at the cost of the overall inquiry. 

Before long, I will post the overview here--now featuring seafood!

Great feedback. Thank you.

________________________________

Levinson, S. C., & Holler, J. (2014). The origin of human multi-modal communication. Phil. Trans. R. Soc. B, 369: 20130302.

Tomasello, M. (2008) The origins of human communication. Cambridge, MA: MIT Press.

___________ (2009). Universal Grammar is Dead. In Evans, N. and Levinson, S. C. The myth of language universals: Language diversity and its importance for cognitive science. Behavioral and brain sciences, 32, 470—471.

:-)  Thanks, Pablo. Sorry for the misspelling of Dobzhansky.

FYI, in attachment the PPT of my other lecture at the conference Human Evolution - Past, Present & Future (London 8-10 May 2013, with David Attenborough & Donald Johanson), on the coastal dispersal & waterside evolution of Pleistocene Homo.

@Michael: Thank you. I'm going over those studies too now, on a course. A very interesting approach, with plenty of compelling data. Apparently, more and more they introduce computational techniques to double-check experiments, and also vice versa. 

@Marc: Thank you for sharing. I'll be using the first one you posted more, though the second one is even more jaw-dropping. How much is known already! There is often whining about lack of evidence on language evolution but how come... I guess the problem is more about the articulation of all the evidences from the myriad subfields involved, which sometimes come othogonal or opposite to each other.

Hi Pablo. This may not be directly relevant to what you are interested in, but I believe that I have a rather novel and comprehensive approach at least to the (gradual) evolution of syntax (grammar). My book on the topic (Evolutionary Syntax) is in production with Oxford University Press, and should be out in June of 2015. In July of 2015 I will be teaching a course "A Program for Evolutionary Syntax" at the Linguistic Society of America (LSA) Summer Institute at the University of Chicago. http://lucian.uchicago.edu/blogs/lsa2015/

My approach reconstructs the previous stages of syntax by using a precise method of reconstruction based on syntactic theory. Even though the syntactic theory adopted for reconstruction is that of Chomsky's (1995) Minimalism, my reconstruction leads to radically different results, to the conclusion that syntax evolved gradually, subject to natural/sexual selection pressures, and that the previous (well-defined) stages of syntax are not only still visible in various grammatical constructions across languages, but that this evolutionary scenario is the best way to explain why syntax is the way it is today, and why languages vary exactly across these dimensions. 

You can find one chapter and contents of the book, as well as some previous papers published on the topic in e.g. the journal Biolinguistics, posted here on ResearchGate. I also have a website which has this information: lprogovac.com.  

Thanks for asking this question.

Thanks, Ljiljana, looks very interesting. No doubt syntax evolved gradually (in small steps) and was subject to natural selection (sex.selection is more about music, I'd think). I'll have a look at your website. May I ask: have you read SC Dik 1997 "The Theory of Functional Grammar" Mouton de Gruyter NY?

Thanks, Marc. I do not discuss Dik's work in the book (I probably should have), but I do discuss e.g. Talmy Givon's work, Bernard Comrie's work, Heine and Kuteva's work, as well as that of many others. I have found that various frameworks and approaches have a lot to offer to resolving the question of language evolution. 

Yes, I agree with you that music and singing were/are most clearly subject to sexual selection, but there are some very specific linguistic data that point to possible sexual selection for the most basic syntactic constructs, which in various languages happen to be rather creative and striking insults (discussed in a joint paper with John L. Locke, "The Urge to Merge," as well as in the book and other papers). In general, I try to relate some specific data and phenomena about language (or syntax) to the postulated stages of the evolution of syntax, and then determine how they would have evolved, rather than have blanket statements about the evolution of syntax in toto. I found that when the specific stages are postulated, a rather clear picture emerges regarding why and how selection would have targeted a more complex grammar, as well as why it had to be in steps. 

Hi Liljana,

In light of the method you used, the results make quite a turn of events indeed. Such a closer look at syntax, in combination with recent evolutionary advances, seems a timely addition to the state of affairs of this topic. Looking forward to reading more!

I think that the Société Linguistique de Paris in 1866 refused to admit any further papers on the subject was a very, very wise decision.

I thought syntax evolved gradually: first isolated words S (word=subject), then SV (+verb), then SVO (transitive verb), etc., as discussed by Dik, but your ideas look also very interesting, thanks a lot, Ljiljana.

My papers (New Scientist, Med.Hypotheses, Hum.Evolution etc., lectures, book (in Dutch) & ebook) are about earlier phases: the elements of speech itself: vowels, consonants, tones, rhythm, dialog & brain structures: early hominoid duetting (cf gibbons), much later supplemented by Homo's early-Pleistocene littoral dispersal along African & Eurasian coasts + seafood collection: voluntary breath-hold diving (breathing control), seafood swallowing instead of biting (consonants), and brain enlargement (DHA etc. in seafood). www.researchgate.net/profile/Marc_Verhaegen independent.academia.edu/marcverhaegen

Thanks, Pablo and Marc! I will look at the profile, too. 

In my reconstruction of syntax in the 2015 book, I was following the reconstruction method somewhat rigidly to prevent myself from coming up with impressionistic conclusions. What did not follow from the method based on syntactic theory, I did not reconstruct. This limits what I could do or say severely, but I thought that it at the same time made the (limited) reconstruction more focused and plausible. I also reconstructed an intransitive stage, but an absolutive-like intransitive stage, which does not distinguish (syntactically) between subjects and objects. Interestingly, modern languages (possibly all) still have such simpler foundational absolutive-like structures, which I discuss. I also have suggestions in the book as to how one can subject these hypotheses to empirical testing, both in neuroscience and genetics. But my model does not allow me to reconstruct the actual word order, so I leave that as an open question, after discussing it to some extent. 

How language evolved to me is the most fascinating question and my hope is that we will look at it from many different angles, and keep the conversation going.  

Thanks, Ljiljana, yes, 150 year ago, the Soc.Ling.Paris was perhaps right in refusing papers on speech/language origins, but that's no the case any more IMO: the whole picture has changed since we know that human ancestors in some phase(s) dived a lot (Max Westenhöfer, Alister hardy, Elaine Morgan etc.): for diving you need a voluntary control of breathing, which is one of the preadaptations of human speech. Later (in the 1990s AFAIR), Roger Crinion found comparative evidence that human oro-pharyngeal innovations are best explained by a shift from biting/chewing (terrestrial foods) to sucking/swallowing (soft & slippery foods): as opposed to chimps, humans have a short mouth, small mouth opening, closed tooth row + more incisiform canines (without dental gaps: allowing suction), smooth & vaulted palate + smooth & globular tongue nicely fitting in the palatal vault, a descended hyoid vs the mandible (a descended larynx vs the hyoid was already seen in all hominoids, presumably for producing loud & varied songs). These innovations probably happened when early-Pleistocene Homo (erectus & relatives) trekked along the African & Eurasian coasts (>1.8 Ma) as far as Algeria (Aïn-Hanech: coastal plain), Georgia (Dmanisi: big river confluence near the Black-Casp.Sea connection), Turkana (where H.erectus appeared together with stingrays) & Java (Mojokerto: river delta + shells & barnacles). There oro-pharyngeal adaptations made it possible to close the oral passage at different places: lips, teeth, palat, uvula (cf consonant production): humans can swallow molluscs & small fishes without biting, even underwater.

Thanks, Marc, your contribution is very, very interesting, indeed. I am really impressed by such a short summary of so much scientific work.

But - concerning our subject - what does it say? As we humans HAVE language there MUST have been an evolution of the anatomy of the human speech organs which made it possible.

Otherwise – maybe mankind would have developed a sign language like ASL?

It seems to me that you could have a look at work in Systemic Linguistics (Halliday, Martin, Matthiessen etc.). Language has co-evolved with the evolution of the brain. Eco social context: increasing complexity of all the physical and social processes shape and define the conditions of human life. The evolution of the meaning potential of the human species involved the human body as signifier and the ecosystem as signified.

https://www.youtube.com/watch?v=nC-blhaIUCk

Your paper is very interesting, indeed, Marc. I also find the following assumption of yours very useful (for the reconstruction of syntax as well)  "that most of the ‘unique’ features of a species (in casu, human speech) consist of more elementary features, which are less unique."  

Keith, when s.b. says "Language has co-evolved with the evolution of the brain" isn't it pure guessing?

Regarding word order, I attach typological (Gell-Mann & Ruhlen, 2011) and experimental (Schouwstra, & de Swart, 2011) approaches, both showing that syntax mirrors human-universal semantic biases. The latter paper, more refinedly, attributes the SOV/SVO alternative to the distinction between extensional and intensional actions.

As for the skepticism on this topic--as of 1866 and on--, I find it very understandable. Impatience is all very natural, and this inquiry has taken the resources of many a lab, for many decades. Yet slow as it is, it may eventually render a quorum-based definition of language, and hence of a great part of cognition, too. That's no trivial enterprise, or quick for that matter. 

Thanks a lot for the 2 very interesting publications you attached, Pablo.   FWIW, my view on word order:  SVO is easiest to understand (what S does works on O: man hit dog, e.g. most pidgins & creoles AFAIK), but +-"stable" languages without much foreign influences tend to have a freer word order (e.g. Latin, Russian), often SOV (man dog hit), or have strict SOV (e.g. Turkish, Japanese): first the players S & O, then V = what happened between S & O (new information is best placed at the end of the sentence).  SOV languages undergoing strong foreign influence seem to revert to SVO, which is easier to understand for foreigners (Anglo-Saxon-Frisian-Danish SOV > English SVO, Latin SOV + Franconian superstrate > French SVO). Dutch & German are between SVO & SOV, they're "conjugated verb 2d" languages (topic first, then the conjugated verb, but the infinitive or past participle at the end: Tv(S)O(V)): they illustrate how languages can evolve from SVO to SOV (Medieval Dutch was apparently more SVO than present-day Dutch) or perhaps v.v.  This is a very schematic view, of course, in reality there are a lot more influences on word order, e.g. conjugation, declination, pre/postpositions etc.

Klaus, thanks a lot.  Some thought about language & brain evolution:  when we compare human & chimp brains, there are several differences.  Our olfactory bulb is only 40 % of theirs.  Our primary sensory (Areas 3, 1 & 2) motor (Area 4) cortices are about 3/2 of theirs, although the areas for our foot-leg are smaller than in chimps, but our finger-arm areas are larger, and our mouth-throat areas are much larger than theirs. We have areas of Broca (coordinates Area 4 of mouth-throat) & Wernicke (integrates information from Areas 3, 2 & 1 of mouth-throat & from the auditive areas), which chimps seem to lack (except rudimentary), and we have enormously enlarged association areas. Broca coordinates mouth-throat muscles, whether for breath-hold diving, suction/swallowing, singing or/and speaking.  Wernicke is involved in interpreting sounds, music or words. Broca & then Wernicke (interconnected by the arcuate fascicle) might have been the first enlarged association areas, and have become essential for speaking.  Our brain changes mirror (co-evolve with) the anatomical changes of Homo vs apes-australopiths, esp. laryngeal & hyoidal descent & oro-pharyngeal changes, but it's difficult to disentangle the different possible functions: breathing, singing, swallowing & speaking, e.g. see 2 papers I wrote in 1995 (largely overlapping): "Aquatic ape theory, speech origins, and brain differences with apes and monkeys" Med.Hypoth.44:409-413 & "Aquatic ape theory, the brain cortex, and language origins" ReVision 18:34-38.

Marc, you wrote: "when we compare human & chimp brains, there are several differences. Our olfactory bulb is only 40 % of theirs." When we look at the reports on wolf children we can read that their olfactory and other senses (hearing, seeing) were much more developed than that of normal human beings. The question is whether the measures say anything about the functions.

Yes, Klaus, they do, although the brain is more plastic than was thought, e.g. musicians have larger areas insulas (areas where our brain analyses sounds), piano players have larger Areas 4 for the fingers, and London taxi drivers appear to have larger hippocampi (memory). But humans not only have much smaller olfactory bulbs than chimp & certainly baboons, but also have a lot of olfactory genes inactivated (cf our ancestors' littoral past), and even intense training can't help this much.

If I can go back to word order, I can offer a very different view...

As Gell-Manna and Ruhlen (2011) note, the reconstruction of an SOV stage “need not necessarily refer all of the way back to the time when behaviorally modern humans emerged and peopled the Old World. There could have been a “bottleneck” effect at a much later time, with a single language spoken then being ancestral to all or most attested languages.”

If my syntactic reconstruction is on the right track, then the initial stages of syntax were intransitive and absolutive-like, without a subject/verb distinction, and as such could not have had either SOV or SVO structures, or any other type of transitive structure. Instead, this would have been a two-word stage with e.g. a verb and one noun (VN or NV) or perhaps, if this single argument is taken to be a subject regardless of its thematic/semantic role, then we can say the order was either VS or SV. But no objects. Sort of what happens with intransitive absolutive structures in absolutive languages, but also in certain constructions in nom-acc languages (e.g. middles; certain compounds, etc.). One example that can illustrate this stage would be VerbNoun compounds like: scare-crow, kill-joy, pick-pocket (where the noun seems object-like: one who kills joy), and rattle-snake, cry-baby, busy-body, stink-bug (where the noun seems subject-like: snake that rattles), and then e.g. dare-devil (where it seems it can go either way: one who dares devil or devil who dares; native speakers in fact disagree re which one of the two is right). Notice that the syntax/structure of these compounds is rigid and very constrained (two-slot grammar, with the verb first, and the noun second, regardless of its role to the verb). If language started in a comparable fashion, from the very beginning, there would have been some autonomy of syntax, that is, the syntax would constrain what and how much we can express in a single utterance. Even though semantically we can perceive an event as involving an agent, a patient/theme, an instrument, a location, a benefactor, if the syntax mold had only two slots, then one could express only one of these roles, leaving the rest up to pragmatics. Just like with the compounds above. Note that Givón (1979) proposed that there was a presyntactic stage which had a “low noun to verb ratio.” In my analysis, this was a syntactic stage, although of the simplest possible kind.

In this view, this is not to say that semantics and pragmatics did not play a role, and that Agent First was not relevant, but it is to say that these principles did not determine the outcome of syntax, or the ordering. Rather, from the very beginning, continuing to this day, there has been a tension between syntax (and the number of slots it provides in any given construction), and the semantic/pragmatic possibilities. That seems more consistent with the variety of word order “solutions” across languages, and within single languages as well.

Furthermore, if the initial stage of language was this absolutive-like (VS or SV) syntax, then transitivity could have arisen via parallel paratactic concatenations of the kind: Woman push, man fall, as have been reported for the early stages of the emerging sign languages, and as also attested with serial verb constructions available in various languages, either in that order, or in the Woman push-fall man order. What they have in common is one argument per verb (intransitivity). If this is true, then Agent-First as discussed by many can in fact reduce to Cause-First, which is a more comprehensive principle, and which operates across languages in paratactic constructions such as: Nothing ventured, nothing gained. No justice, no peace. No pain, no gain. Easy come, easy go. These do not involve agents or transitivity, but the cause has to precede, as we cannot reverse the order. If so, then Agent First in e.g. Woman push man fall can be a consequence of the more general Cause First strategy.

In any event, I am trying to say that my reconstruction of syntax is not consistent with the human proto-language starting with some kind of natural-seeming transitive word order, whether SOV or SVO. I offer many reasons for that in my book, including an astonishing variation across languages, and within languages, regarding the expression of transitivity. I have attached one chapter of the book that outlines the basic arguments.

The fact of human syntactic variation in basic word order makes it seem completely unlikely to me that there is any innate component to that level of syntax. I dont see why one would even assume that there was ever a single hardwired basic word order. I dont see how that hypothesis fits with either what we know of language or what we know of the brain. 

Hi, Magnus. Thank you for your view. I agree with your bottomline of language evolution, yet I do think there are arguments one could initially draw in favour of a native syntactic system. There certainly exist some innate tendencies in humans, readily visible in newborns at a neurological and at a psychological level; think of the senses and some recognition biases, feeding behaviour, attention calls such as crying...

Those early constraints could make ground for nativist arguments of language evolution--that is, if only language were quite as primarily fundamental for survival, so recent in phylogeny, and identical across cultures, too. Apparently, it is not. 

Besides, would syntactic structures have had enough time to get hardwired in the genes? Much of the evidence on genetics says no. Opposingly, the genes we do have are for communication more generally: otherwise, how could mice and birds have the same Foxp2 as we do?

Yes, I agree. From my perspective all of the syntactic universals we see are more likely to be due to semiotic principles of information processing arising from a general cognitive infrastructure of symbolic interpretation, than to there being any hardwired syntax. So to address your initial question the place where I would look for alternative accounts of the emergence of language in human evolution would be in the field of biosemiotics and cognitive functional linguistics. I think that there is a need for a future synthesis of semiotic work such as Deacon's symbolic species with functional/cognitive theories of grammar and neuro- and psycholinguistics. Such an avenue is to me much more promising than anything building on any generative approach.

I know of one very interesting piece, by Aaron Stutz, just along those lines (linked). He brings together Deacon's theory with the framework of embodied cognition. 

http://journal.frontiersin.org/Journal/10.3389/fpsyg.2014.00834/abstract

Marc,

18 days ago you wrote: „humans not only have much smaller olfactory bulbs than chimp & certainly baboons, but also have a lot of olfactory genes inactivated (cf our ancestors' littoral past), and even intense training can't help this much.“

I doubt that. Wolf children are said to have developed much more olfactory abilities than ’normal’ ones. And S. L. Rubinstein reports that experienced grinders are able to distinguish spaces of one to two thousandth of a millimetre of width with the naked eye, whereas human beings usually are only able to distinguish spaces until one hundredth of a millimetre.

No biologist doubts that, Klaus. Even if children extensively train their olfaction (e.g. wolf children), they're still incomparably worse than canids. Dogs can follow traces of scents 4 days old. Primates are already microsmatic (no unexpected, see their arboreal life, where scents are high in the wind), and anthropoids don't even have a rhinarium, but hominoids have even poorer olfaction than monkeys, and humans have still poorer olfaction than apes. Some patients don't smell at all, but are unaware of this until it's tested. In dogs, olfaction is extremely important, but human can miss this sense without serious problems. Sheep & dogs have 200 million of olfactory cells, rabbits 50, humans 6 mill., and 2/3 of our olfactory genes are inactivated, in apes 1/6 (Y.Gilad cs 2003 PNAS 100:3324). Other microosmatic or even anosmatic mammals are Cetacea. Our atrophied sense of smell is in line with early-Pleistocene Homo's coastal dispersal (S.Munro 2010 Molluscs as ecological indicators in palaeoanthropological contexts, PhD thesis Austr Nat Univ Canberra). It disproves the old Man the Hunter & open plain- & endurance-running ideas, e.g. google "original econiche Homo" table 4.

The aquatic account of human evolution scenario is really an entirely different question from this. Oughtn't it to have its own discussion thread? 

The flaw in your argument is of course that when olfaction already becomes less important from monkeys to apes without any necessity for "coastal dispersal", then obviously the same development from apes to humans does not need to be ascribed to that hypothesis. The standard account is that there is a shift away from olfaction and towards dedicating resources to vision in the entire primate line having to do with the development of depth vision and the forward movement of the eyes. The high reliance on vision gradually makes the sense of smell less useful, and the neural resources are taken over by other senses.

From what I have read of the aquatic theory it is all just as badly argued and simply consists in taking any changing trait as evidence of water. 

Just briefly responding to the thread about syntax:

Magnus and Pablo,

I don't think one should discard approaches to language evolution just because they rely on this theory or another. Instead, we should judge the proposals by how good they are at advancing concrete and testable hypotheses, as well as how good they are at moving toward a  synthesis with other relevant disciplines. In fact, for us to figure out this incredibly difficult question, it will probably take insights from all the approaches, and all the angles.  

Magnus,

1) Arboreal mammals typically have reduced olfaction vs terrestrial mammals, but humans have moreover strongly reduced ofaction vs apes. This fact alone disproves the now obsolete savanna idea that H.erectus ran over the plains after antelopes. It confirms Stephen Munro's "littoral dispersal model" (2010 anthrop.thesis) that Pleistocene Homo followed coasts & rivers as far as Pakefield in England, the Cape & Flores island, beach-combing, diving & wading bipedally for littoral, shallow aquatic & waterside foods. Stephen is the one who discovered the shell engraving: no doubt you have have read the recent paper J.Joordens cs S.Munro cs 2014 "Homo erectus at Trinil on Java used shells for tool production and engraving" Nature doi 10.1038/nature1396. If our ancestors had ever run over open plains as popular accounts of human evolution keep assuming without evidence (google "original econiche Homo"), they had had improved olfaction, but we see olfactory atrophy in humans. From my letter to Nature 325:305-6: "Humans have thermo-insulative subcutaneous fat layers, which are never seen in savanna mammals. We have a water- & sodium-wasting cooling system of abundant sweat glands, unfit for a dry environment (where water & sodium are scarce). Our maximal urine concentration is too low for a savanna-dwelling mammal. We need more water than other primates, and have to drink more often than savanna inhabitants, yet we cannot drink large quantities at a time."

2) I have no idea what you have read on the littoral theory, but apparently it was outdated, please see the attachment from the proceedings (Hum.Evol.) of the 2013 symposium on human waterside evolution in London with David Attenborough & Donal Johanson.

3) In this thread we're discussing language evolution. Human language is based on (a) Miocene hominoid song (cf chimp & gorilla hoots & gibbon loud & varied tones, rhythm & duet) + (partly) seafood consumption when Pleistocene Homo followed the coasts, see above: (b) voluntary breathing control for littoral diving, (c) swallowing/sucking seafood instead of biting/chewing antelope, (d) brain-specific nutrients in seafood (esp.DHA).   (a-d) are the preadaptations for human language: for word formation, see e.g. Nowak in PNAS, and for sentence formation, see e.g. Ljiljana Progovac above.

I agree, Ljiljana. As I see it, syntax is no less certain, or tentative, than any other scopes in this inquiry. Only, in the mind of many, syntax will evoke a strong association to the good old generativism, no matter what. Now I think your book is going to help to solve that. 

Marc Verhaegen: To even state that any single change in a trait "disproves" any evolutionary hypothesis about the environment of adaptation is so weird it makes my head spin. I cant even wrap my head around how decreasing importance of olfaction can possibly be considered decisive evidence for any hypothesis of human evolution whatsoever. Surely that could be compatible with any number of possible environmental niches.

I have read some recent stuff about the AAT (including Philip Tobias endorsement), but really it strikes me as unconvincing in the extreme. It may look good in a simple dichotomy between savannah and waterside - but luckily hardly anyone in paleoanthropology thinks in such simplistic terms. Yes humans need to reside close to water. But it it a gigantic leap from that to deriving all other adaptations from that single fact. Much more so to propose that water provided the necessary pre-adaptations for language, in spite of the mountains of evidence demonstrating other factors being much more decisive in that process. 

I may change my mind when your conference proceedings start getting cited in mainstream publications. I would be interested in reading Langdon. Johanson and Stanfords contributions, to hear their evaluations, but I cant find them online. Apart from them and a few others the line-up for the conference looks mostly like the longstanding AAT clique. Certainly not an event that looks like it would have caused a major paradigm shift. 

Liljana, I also agree with your statement that all venues are useful. I dont have a problem with syntax at all, I just dont believe that a formalist perspective on syntax (generativist or otherwise) is likely to prove useful in understanding language evolution (because I dont believe syntax is evolutionary fixed). I may have misunderstood your general approach having only read your posts in this thread, if that is the case I apologize.  I will nonetheless look forward to reading your book which looks extremely interesting and certain to provide valuable insights. 

Magnus, inform properly (i.e. read the recent literature) instead of refusing to read our papers, and think a bit for yourself instead of blindly following the conservative clique who still believes that our Pleistocene ancestors ran after antelopes on the African plains.

As for Langdon (a very nice man, I met him a few times), see the attachment of Mario Vaneechoutte (I co-authored, but I a minor role in it).  Algis Kuliukas also has written an answer to Langdon's papers IIRC.

Thanks, Pablo and Magnus, for putting trust in my book. All I ask is that people read it with an open mind, both those working within Minimalism, and those working within other frameworks. 

As Pablo said, Minimalist syntax is no better or worse than any other field. It has some truly great insights, and some very unconvincing claims, and a lot of controversy and disagreement within the field. For example, the most recent claims that syntax is somehow perfect and optimal (like a snowflake) is rather unconvincing to me, given all the variation and quirks across languages and within languages. And because such claims are formulated in such a way that they cannot be falsified. On the other hand, this approach has produced an amazing amount of data, descriptions, and precise analyses for a variety of languages, and thus provides good tools for analyzing sytntactic variation across languages, as well as within languages. 

Its analysis of syntactic structures as hierarchical layers provides a particularly suitable tool for reconstructing evolutionary stages of syntax, and for corroborating such stages with a precise analysis of structures of various degrees of complexity found across modern languages. It allowed me to reconstruct the stages of proto-syntax by relying on a theoretical framework, rather than on impressionistic claims or common sense. Having said that, I found it encouraging that once I reconstructed syntax down to the simplest stage, this reconstruction now also makes perfect common sense.