Вълшебни конструкции, мониторци и сегагейци.
Отиди на страница:
| Това в картинката на Звездников е наденица. Появиха се една поредица подобни, по повод доклад на WWF на тема как богатите изядоха храната на бедните. Извинете за отклонението. |
| tupakmango, УВАЖАВАМ(в истинския смисъл на думата) всичките си ФОРУМНИ другарчета! Без майтап!  . P.S. Да не забравя: - Никога не съм стрелял с "кокички от вишни"(в някои версии - коктейлни череши) м/у рогите на Вежливия Лос! Е...че му порастнала нек'ва си там японска вишна или череша...не е моя заслуга!  Редактирано от - OLDMAD на 14/05/2006 г/ 20:35:01 |
| Портокалената Принцеса или Сингулярността на удоволствието През една майска вечер преди около век, класикът на американската антропология Франц Боаз стоеше пред вигвама си любимата си поза лотус, a взорът му блуждаеше над силуета на скалистите планини в южна калифорния. Не можеше да реши дали този силует прилича на мръснозелено морско водорасло, дали на профила на Карл Маркс (ако го погледнаш с глава наклонена наляво), дали на млечните жлези (всъщност цици, но политическата коректност не му позволяваше да ги назове така) на дъщерята на местния вожд, Портокалената Принцеса Без Възраст, стара като нощта и млада като луната, с които се бе запонал лично на празника на пъпчивото куче предишната нощ. Тъкмо беше решил, че силуета на Скалистите планини всъщност прилича на еректиралия пенис на самия вожд, Царевичен Кочан (с когото също се бе запознал лично) , когато му мина мислъта, че характера на сравненията се дължи на халюциногенния кактус, който беше дъвкал през празнеството на пъпчивото куче, организирано в негова чест, на което той, Франц Боаз, беше получил своето индианско име, Учтивия Козел. Можеше да бъде и по зле, мислеше си Франц, например Метилявата Овца или Шугавият Пръч, както се беше случило с други етнографи от други епохи, изследващи оргиастичните ритуали на племената сегахо и мониторези. Франц Боаз беше от школата на лингвиста-антрополог, Saussure, чието име име американците произнасяха като Соссьор, още едио име, което източноевропейските философи бяха проспали в полувековния си сън, тъй като при тях за авторитет в лингвистиката се считаше - познайте от три пъти - а. Маркс б. Карл-Маркс с. Маркс-Енгелс. (И трите пъти сте прави). Теорията на соссьор гласеше, че частите на речта нямат отношение и смъсъл нито за говорещия, нито за адресата му, смисъл има само цялото изречение. Сьосьор беше направил простичкото наблюдение, че едно изречение, за да бъде разбирамемо, трябва да се подчинява на определена организация - да има сказуемо, подлог връзка със сказуемото, обстоятелствено допълнение и т.н., и всички изречения, произнесени на всички европейски езици, се подчиняват на тази организация. Това, което сосьор открил и с което хвърлил във възторг социоантрополозите, било, че всяко изречение притежава структура. И че структурата на изречението няма никакво отношение към смисъла на изречението, нито към автора си. Антрополозите като Боаз изпаднали във възторг, защото със структурната си лингвистика Сосьор изградил мост между езикова структура и обществена структура, представил на вниманието им първата система, която принадлежи на нематериалното, няма физическо тяло, не може дабъде пипната, помирисана и т.н. и в същото време е напълно реална - граматиката, и поставил началото на структурализма. Първия нематериален обект, който доказано съществува, и по-важното - за разлика от останалите феномени на нематериалния (т.е. социо-културния) свят, е измеряем, изчисляем и подлежащ на описване и изучаване. Това, което леко безпокояло Соссьор, е че структурата на изречението, за разлика от самото изречение, НЯМА АВТОР. По късно Ноам Чомски си спечели медалите като доказа, наличието на граматически инстинкт у всяко дете още в предграматическата фаза на развитието си, т.е. умението за организирането на думите в структура май се наследява. Естествено, за да се развие е необходим контакт със себеподобни, т.е. социално канализиране на умението. Редактирано от - Rednik South O.Z. на 15/05/2006 г/ 18:42:09 |
| Разбира се, половин век по-късно Леви-Строс вкара гол от центъра като обяви, че митовете всъщност са езиков феномен, разположен, за разлика от синхронизма на изреченията, диахроничко - първата версия на мита отговаря на сказуемото, следващата му версия 100 години по късно отговаря на подлога, птретата добявя още една думичка и т.н., тоест мита е език от втори порядък, дето уважаемата Геновева го нарича метаезик, където една генерация произнася първата дума, следващата втората, алализатора на мита вместо да анализира несъзнателно добавя третата и т.н. до безкрайност, като единичната, сингулярна версия на конкретен мит има точнотолкова смисъл, колкото подлога в език от първи порядък - т.е. никакво. Ептен по късно пък Деррида направо счупи гредитеи на леви стросс, и на соссьор, ма тва не е по темата. Енивейз. Това, което безпокояло Боаз, бил факта, че в езикът на индианците сегахо и мониторахо липсвали времената и падежите, но за сметка на това имали 6 (шест ) парчета от непознатата в европейските езици Граматическа фигура Х, която определяла местоположението на говорящия в пространството - дали описвания в изречението обект се намира вдясно от лъка му, дали е вляво от копието му, дали му е зад ..гърба...или пред..оная работа, дали е под нивото на мокасините или над нивото на скалпа, и Боаз се чудел как да отрази тази граматична форма в речника, който пишел, за да бъде разбран от европейците.Освен това Боаз не бил убеден дали метафорите, с които замества действителността, са разбираеми и усвояеми от индианците и Портокалената Принцеса, въпреки че са произнесени на същия език, защото Боаз не е сигурен дали познава и разбира граматическите структури на езика на Сегахо-Мониорезите. Особенно го измъчват грамтическите формулировки за местоположение на говорещия по отношение на слънцето и отсъствието на времена и други форми, което прави ебах маа му да звучи като еба мама ти, и освен това никога не може да разбере кога го ебат, кога го псуват и кога му рецитират хайку. Като резултат Боаз стои в поза лотус пред вигвама си в скалистите планини и си мисли че удоволствието от общуването никога не може да бъде взаимно, защото в природата на общуването е да бъде невъзможен трансфера на информация от езика на сегахо на езика мониторези, и вайс верса и общуването е занимание самотно. |
| Добър микс. Сосюр се пише на български - Saussure. И сказуемото се съгласува с Подлога, не с Предлога. А къде е оня индианец, дето настъпил мотиката в тъмното?  |
| Има едно доминиращо обвинение срещу Фуко, че на теорията му за Властта, Подчинението и Съпротивата и липсва достоверността на Марксизма, защото маркса сочи източник на властта - това е доминиращата обществена класа, центъра, слънцето, докато според Фуко Властта няма локус, т.е. център, също като гравитацията - ако слънцето изчезне законът на гравитацията ще си продължъ да действа и ще пренареди планетите в друга конфигурация. Имам една малка работа по въпроса за властта и съпротивата, която поствам тук. Нямам я на български. ######### In search of Monsieur Bricoleur: Power, Subject and Resistance “Philosophy turns into the empirical science of man, of all of what can become for man the experiential object of his technology, the technology by which he establishes himself in the world by working on it in the manifold modes of making and shaping......No prophecy is necessary to recognize that the sciences now establishing themselves will soon be determined and steered by the new fundamental science which is called cybernetics. This science corresponds to the determination of man as an acting social being” M. Heidegger, 1974, Pp 8-12 Many times, including in the classrooms of Duke University, the theory of M. Foucault has been criticized for not exposing a subject of power and a conclusion has been drawn that for the above reason Foucault’s theory lacks credibility. What we ignore in our quest to verify or deny Foucault’s theory by identifying a possible subject of power and possible ways of resistance is that the natural, pre-cybernetic systems, which are self-organized, lack natural center (the physical universe, the ecosystems, the language systems are all examples of systems, functioning without a central subject, or an author, unless we want to go into the field of theology) and these non-artificial systems, unlike Microsoft Windows, do not have a system engineer as a cause of their existence. In addition, they certainly do not have a central subject, which defines the rules or distributes power, again, unless we want to depart the field of social sciences. The goal of this essay is to prove that power could not possibly have a subject, to show how the lack of subject affect the modalities for resistance and to extract Foucault's notion of power from the realm of his book "History of sexuality" - a realm, where the society as a system do not consists an author and power that rules does not have a subject. We live in an epoch, where artificial cybernetic systems emerge and shift the patterns of our living every day, exactly as Heidegger professed. Far from substituting the philosophy as a science for the human as a social being, (M. Heidegger, 1974, Pp 8-12), a cybernetic by its nature notion of the society as a self-regulating system has been introduced by a number of social theorists, from the Neo-Marxists trough Derrida and Michelle Foucault. Engaged in a permanent encounter with an endless constellation of human-made cybernetic systems, resulting from the advancement of engineering science, somehow we are tricked to neglect the fact that we co-exist in a system that timely precedes the human-made cybernetic systems. This system is the society. Based on the every day’s experience, our senses mislead us to believe that every system has an author, its system architect, which masters the structure of the system and its order. We look for the singular cause of the system’s existence and its logical center, where the subject of power resides. In the realm of Michelle Foucault, similarly to all natural systems, which are a product of self-organization, the social systems and its subcomponents have no author, or at least, their author is not presented in the system. These systems are organized into a monstrous, ever expanding web of subsystems by a force that Foucault qualifies as power. Foucault identifies power as a "multiplicity of force relations, immanent in the sphere in which they operate and which constitute their own organization", and as a "process, which transforms, strengthens or reverses them"(M. Foucault, 1978, p 92). What is obvious from his definition is that power is neither a structure, nor a component of a structure. Power is a relation, a force relation that exists between the structural components. At the same time, since power constitutes, creates the organization and therefore the structure itself, according to Foucault power is the cause of the existence of the structure. Another feature of power is that it has no fixed place in the structure - the power is a "moving substrate"(M. Foucault, 1978, p 93) of the force relations between the structural components, which also points the extra-structural nature of power. There is no component in the structure, which is a domain of the power - "..it comes from everywhere" (M. Foucault, 1978, p 93). The place of power could not be identified or associated with any particular social component (or construct) - individual, class, strata, gender, family, etc. Still, the power is the organizing force that gives shape and organization - a structure - to the society (“makes it function”, M. Foucault, 1978, p95). The society, being a domain of the power, which comes from everywhere and resides nowhere, is a haunted house. Despite of operating in several distinctive modalities, the technology trough which power manages the structures of sociocultural continuum can be reduced to the use of a single, very powerful instrument for creation, management, proliferation and reproduction of the socio-cultural structures - the discourse. (M. Foucault. 1978, p138, p101). Foucault defines discourse as diverse by its nature elements (some of these components represent the infrastructure and other are discourses themselves), that are “grouped together in an artificial unity” (M. Foucault, 1978, p154), as this “fictitious unity” (M. Foucault, 1978, p154) is used as a “casual principle” (M. Foucault, 1978, p154) for the creation of more discuses and this is how the social structure is created. In the technology of creation of any socio-cultural structure (society, language system, etc.) the discourses function as “tactical elements or blocks operating in the field of force relations” (M. Foucault, 1978, p101). Based on the above-cited passages, we can conclude that the discourses have entirely social character, they are all produced, they do not represent reality, and they are not to be found in nature (“we must not place the sex on the side of reality”, M. Foucault 1978, p157), they are pure social facts, entirely “imaginative“ and still very real, but only in the social universe (M. Foucault, 1978, p78, p154, p156-157). Thus, the social universe along with all its structures is in no means a representation of the reality - it is an entirely different world, where signification does not means representation of the natural world- what the discourses signify are another discourses, which does not necessarily means that the natural world and the social world do not have points, where both worlds meet, interfere, confront, and penetrate each other, which is a very major theme in “History of Sexuality“(M. Foucault, 1978, p138-139, p142-144). ############## Thus the discourses are produced, subjectified by power trough the use diversified components and further used as a building blocks for the production of the next level of discourses - the structural components and the subjects, which are discourses of higher order, since they include a number of discursive substructures as building materials. The power constitutes every subject in the social universe - “To the structure, that constitutes the subject himself, one finds a general form of power, varying in scale alone” (M. Foucault, 1978, p85) and, further, “confronted by a power that is a law, the subject who is constituted as subject - who is “subjected“ - is he who obeys“ (M. Foucault, 1978, p85) . The power is presented in every subject, it constitutes every possible subject in the socio-cultural field, and no subject can escape the omnipresent grip of power. Even the man itself is produced by the discourse-based mechanism of power as a subject in the social field - “The West has submitted generations in order to produce men’s subjection” (M. Foucault, 1978, p60). It seems that in the socio-cultural universe, every subject has been introduced by power, including the man as a social entity. The question that rises when encountering Foucault‘s theory of the power as a primal cause for the existence and the author of the social-cultural system with all its structures and subjects, is what or who is the central subject of power and what is the source of power. With other words, where does power come from? If we accept Foucault description of the world as a system, power could not possibly have a "subject", since power has created all subjects in the society trough different combinations of discursive formations and logically could not precedes its own subject in the casual chain. If we accept that power is the only subjectifier that produces all the subjects, no subject could precedes the power, including the subject of power. Thus, the very expression “subject of power” becomes an oxymoron, something like “wooden iron”. To go further, since the power is the cause of the existence of the socio-cultural universe, its "creator", power precedes in the casual chain the entire social system (all subjects are positioned in the social system, they are not natural facts). In addition, since power precedes the social system, which has created, by nature power appears to be external to the system. And since it is the external cause of the system, the true nature of power is transcendental, does not belong to the social universe and its substance could not possibly be studied - it is unintelligible in socio-cultural terms, since it is transcendental to the social milieu. But if the true nature of power is transcendental, power that operates the social universe becomes only a representation itself, not the thing, that is the external cause of the system, hence, what we perceive as a “power” (or what Foucault perceives as “power”) is only a representation of an x-force, that is positioned outside the sociocultural system, in the external world - out of the system, which could mean that this x-force resides either in the theological world or in the physical(natural) world, but not in the social one with which Foucault deals. Therefore, when Foucault concludes that power has a meta-discursive character and is omni presented in all discourses, he is somehow imprecise - power is only relatively meta-discursive, since it is just another "representation", not a "presence". Ad since power is only a representation of a "thing", not the thing itself, "power" it is just another discourse, which is a failure of the formal logic, but that is the way it seems to be and the failure of the formal logic when dealing with the phenomenon of power again testifies for the transcendental, non-human, hence non-intelligible origin of power. If the source of the power is transcendental, this source could not be a subject of knowledge for the subjects that reside within the system - it could not be possibly subjectified. In the light of the above conclusion, the question of resistance against the power becomes a very interesting one. The ultimate thing that we could do in order to break the power matrix is to eliminate and un-do all subjects, including our very self and all socio-cultural structures, since they have been created by power and to wait for the reality to arrive. My suspicion is that we may not actually like it, if all we know about this reality is discursive, hence- provisional, so I personally find the above way of resistance rather risky and unpredictable, although this is the most secure one. However, since we, as a system residents, could not have a knowledge of what resides outside the system, there is no guarantee that even if we destroy all subjects and structures in the universe, the power will not re-emerge and re-built everything that we have destroyed, since we can not hurt the cause itself, but only its representation in the system. Another, probably more reasonable way of resistance against the power is to un-do, un-learn and de-construct as many discourses as we can, without becoming a starting point of new discourses -the way E. Said tried to do it with the orientalism. As a deviation of the above way we can reduce as many structures of power as we can, living untouched only the most necessary for the existence of the society, and to prevent the emergence of additional power structures. The third possible way of resistance against power is the way of Monsieur Bricoleur and Derrida, by putting the structural components into play, to see what is going to happen and to enjoy it. I personally prefer the third way. ############ 1. Heidegger, M. Zur Sache des Denkens (1962-64). Translated as The End of Philosophy and the Task of Thinking by Joan Stambaugh (New York: Harper & Row, 1972), pp 8-12 2.Foucault, M., The History of Sexuality, an Introduction, Volume I (1978). Translated from French by Robert Hurley (New York: Randome House, Inc., 1978) pp 60, 78, 85, 92-93, 95, 101, 138-139, 142-144, 154, 156-157 |
If the source of the power is transcendental, this source could not be a subject of knowledge for the subjects that reside within the system - it could not be possibly subjectified. И кво от това? А това пък го не разбрах non-human, hence non-intelligible origin of power ???? _____________________________________ Б.ВЕЛЧЕВ: ...не си струва да опитваме да се справим с корупцията... |
| Rednik, бе не успях да го прочета - губи ми се нещо мисълта. Има ли някъде абстракт на това, така някак си, кондензирано, само есенцията? |
| Здравей Боза, извини ме, че сега не мога да направя каквото искаш. Ако още ти е интересно, следващия път. Всъщност пуснатото е ексерт - т.е. най важните части, не цялото, нещо като разширен абстракт. Поздрави и всичко хубаво. |
ЛелеееЕЕeee, тука продължава конкурса по осторумие на произволна ироническа тема.... Тц, тц.... Двубой в свободен стил! Хайде наш'те!!!!  |
| Прав си - This it is, and nothing more _______________________ Можеш ли да докажеш на човек, видял розови слонове, че те не съществуват? |
| Новина за българските пророци на глобализацията: Миналата седмица сенатът на бившата цитатела на глобализма като визия за бъдещето, уесе-то, гласува закон за обявяването на английския език за официален и за елиминиране на испанския като език на държавните институции. Америка официално се обяви за едноезична държава, използването на испански извън частни разговори се инкриминира, т.е. приравнява към престъпление. Добре дошли в прекрасния нов свят на неонационалните държави и националния ренесанс. Продължавайте да плувате пампор по въпросите на малцинствата, мултикултурализми, колективни права, майчини езици и тактически концепции отпреди 20 години, докато светът се отдалечава в противоположна посока. Берлинската стена се премести от берлин в аризона и бетооно-телените колони и мрежи са като призракът на Стената от студената война и като овеществена метафора на великата китайска стена, която някога регулираше преселенията на номадите и пазеше китай от другостта на другите. Хората строят берлниски страни и институционализират езици, а ние се водим от по същество тактически прийоми и крещим да здравствует глобализация. Как беше вица - Лявото крило гори -тра-лал-ала-ла -дясното крило гори---тра-ла-ла-ла-ла... Да видим европа къде ще си построи китайската стена и от коя страна на стената ще попадне Бъгария....наздраве. |
| Редник, мисля че прекаляваш? Приемането на закон за английския език като основен, няма нищо с "елиминиране на испанския като език на държавните институции". -Изобщо! Не съм чул нищо подобно и не вярвам в никакъв случай на такива неща - макар, че ако е верно съм първия дето ще го приветства. Невъзможнo е да стане! Това са свободни интерпретации. А за стените си още по-далеч от истината... ~Всичко си остава както си е било. Приказки, закони? |
| Здрасти Доктор Звездников, първо да ти пожелая здраве и дълъг живот, а останалото, както казват комшиите - башка. По- късно ще ти отговоря, законът за институционализация на английския го гласуваха в четвъртък, а стената, вярвам, ще имаш удоволствието да я видиш с очите си, и нейното построяване гласуваха в същия ден, голям завой е това, нищо няма да като преди, като се разигра мечката на промяната преди 15-тина години и още не иска да спре. Не сме на едно мнение с тебе горе долу по нищо, но, моите уважения и най добри пожелания! |
Какъв е тоя английски, Южен. Това не е език, а кражба. Аз си се оправям чудесно с френския, окраден от английския в 70 процента /казват, от "олд френч"/, когато англичаните били прости селяни, нечленоразделни, а французите ги учели на изисканост и на език - тия пейзани - англичаните. Ще ме прощаваш за откровеността. Редактирано от - Сибила на 31/05/2006 г/ 17:33:04 |
| .... добре изучихме франсетата, че те после да изучат ингилизите ..сал мадриТското чучело не успееме изучим |
| Преди време форумецът Тормозчиян постна съобщение за генетични доказателства за интербрединг (хибридизация) между предшествениците на хората (Тумай) (т.е. хуманоидите) отпреди 5.4. и предшествениците на шимпанзетата. Направих си труда да потърся оригиналната публикация, която поствам тук. Хипотезите, доколкото мога да преценя, са брилянтни. Коментар към постинга - когато мога. Още по куриозни биха били алтернативните обяснения.... Публикацията е от платената част на нейчър, където е публикувана оригинално, и не е със свободен достъп. За да избегна проблем с авторските права, съкращавам 1/10 от нея, за да мине в графа разширен цитат. Ако има проблем все пак, бот да трие, щото проблема ще е общ. Вскички права принадлежат на долуупоменатите. ******** Article Nature advance online publication 17 May 2006 | doi:10.1038/nature04789; Received 5 November 2005; Accepted 7 April 2006; Published online 17 May 2006 Genetic evidence for complex speciation of humans and chimpanzees Nick Patterson1, Daniel J. Richter1, Sante Gnerre1, Eric S. Lander1, 2 and David Reich1, 3 Abstract The genetic divergence time between two species varies substantially across the genome, conveying important information about the timing and process of speciation. Here we develop a framework for studying this variation and apply it to about 20 million base pairs of aligned sequence from humans, chimpanzees, gorillas and more distantly related primates. Human–chimpanzee genetic divergence varies from less than 84% to more than 147% of the average, a range of more than 4 million years. Our analysis also shows that human–chimpanzee speciation occurred less than 6.3 million years ago and probably more recently, conflicting with some interpretations of ancient fossils. Most strikingly, chromosome X shows an extremely young genetic divergence time, close to the genome minimum along nearly its entire length. These unexpected features would be explained if the human and chimpanzee lineages initially diverged, then later exchanged genes before separating permanently. The genetic divergence between two species (the proportion of nucleotides differing between representative individuals of the two species) can be converted into a divergence time in terms of millions of years, provided that differences between genomes have accumulated at a constant rate as a result of new mutations1, 2. The average genetic divergence, genome, is sometimes used to estimate the speciation time, species. However, (x), the genetic divergence at any position x, fluctuates across the genome and is everywhere larger3 than species (Fig. 1a, and Supplementary Note 1). Thus, its average genome necessarily exceeds species. a, The genetic divergence time between two species, (x), varies across the genome and is always greater than or equal to the speciation time, species, which is the time of last gene flow between the species' ancestors. The average genetic divergence (genome) thus always exceeds species. b, The historical relationships of the species included in this study, along with relationships to various fossils (adapted from ref. 43). The relative lengths of the branches can be estimated from the data by the number of divergent sites of each type. c, d, Genealogical relationships are not always the same as the species relationships (grey), because humans and chimpanzees can sometimes share a common ancestor that is older than the gorilla speciation (greater than ). For example, although humans and chimpanzees are most closely related in most sections of the genome, there are regions in which chimpanzees and gorilla are most closely related4–6 (producing 'CG' sites, c), or in which humans and gorilla are most closely related ('HG' sites, d). e, A revised model that could explain our data is that the first hominins became isolated from chimpanzee ancestors more than 6.3 Myr ago, but then hybridized back to the chimpanzee lineage. This could explain the great variation in divergence time across the genome, with humans and chimpanzees sharing a common ancestor around the time of hybridization in some regions (blue line) and before the initial speciation in others (red line). A model with chimpanzee ancestors as the hybrids is equally consistent with the data. Although chimpanzees are our closest relatives, there are many loci at which humans and gorillas (or chimpanzees and gorillas) are the most closely related4, 5; we estimate that this is the case over about 18–29% of the genome (Supplementary Note 2). In such places, the genetic divergence time of human and chimpanzee must precede gorilla speciation (Fig. 1b–d). Thus, humans and chimpanzees show large variation in (x), making them an excellent system in which to explore how taking into account the difference between genome and species can affect inferences about history. Previous genetic analyses of great apes studied small data sets (the largest was about 25 kilobases (kb))5 and suggested that the time since average genetic divergence of humans and chimpanzee genes is much greater than the time of speciation4, 6, 7, 8. However, they produced inconsistent estimates of ancient diversity owing to small data sizes, ignoring the effects of recurrent mutation9, and simplifying assumptions about the demography of ancient populations10. Genome comparisons of five primates To create a much larger data set, we generated shotgun sequence from the gorilla (about 115, 000 fragments comprising about 87 megabases (Mb)), and compared it against the human11 and chimpanzee genomes12 and the unpublished sequence of orangutan and macaque (Methods). We overlapped the sequences, producing four-species human–chimpanzee–gorilla–macaque (HCGM) alignment (18.3 Mb) and five-species human–chimpanzee–gorilla–orangutan–macaque (HCGOM) alignment (9.3 Mb) (Supplementary Tables 1 and 2). We also studied 1.2 Mb (ref. 13) from contiguous regions of chromosomes 7 and X (Methods). Altogether, these data represent more than 800-fold more aligned bases than the largest data set previously available5, and enough data to compare chromosome X with the autosomes. To analyse these data we identified all 'divergent sites', places at which two alternative alleles were observed across the aligned sequences of the species. We eliminated sites in hypermutable CpG dinucleotides11, and those not flanked by at least one base of completely conserved sequence (our qualitative results are unaffected by these filters; Supplementary Table 3 and Supplementary Fig. 1). This produced 858, 941 divergent sites for the HCGM shotgun data, 498, 771 for HCGOM shotgun data, and 78, 290 for the contiguous data. We categorized the divergent sites according to how they partitioned the species (Table 1). In the four-species HCGM alignment, there are seven possible partitions: four in which one species differs from the other three (denoted H, C, G and M) and three in which two species differ from the other two (denoted HC, HG and CG). If all divergent sites were due to single historical mutations, the proportion of each class, nH:nC:nG:nHC:nHG:nCG:nM, would be strictly proportional to relative 'branch lengths' of the genealogical tree—that is, the elapsed time on each branch averaged across the genome (assuming that mutations have accumulated at a constant rate over time1, 2) (Fig. 1b–d). However, the correspondence is not exact because some sites are due to more than one mutation9. Recurrent mutation has a particularly distorting effect on short branches. Because HG and CG sites occur rarely and can easily be generated by recurrent mutation, branch lengths tHG and tCG would be overestimated without this correction. By using five-species alignment data (HCGOM), we were able to estimate the effect of recurrent mutation on branch length estimates, controlling for the biases it introduces. In particular, the HCGOM data revealed six classes of divergent site that could not be due to single historical mutations (Table 1). An example is HO, which clusters together human and orangutan (Supplementary Table 4). We developed two methods that use the rates of these sites to correct the estimates of branch lengths for the impact of recurrent mutations (Supplementary Methods and Supplementary Note 3). Our analysis indicates that past studies that failed to correct for recurrent mutation5, 6, 7, 8 obtained roughly twofold higher estimates of the rates of HG and CG sites (Supplementary Table 5). Thus, these studies were biased towards overestimating the proportion of the genome in which humans and chimpanzees are not most closely related5, 6, 7. Large variation in divergence time across genome We used a straightforward approach to study variation in (x), avoiding the assumptions about the demography of ancient populations from previous studies5, 6, 7, 8. We selected subsets of the genome in which we proposed that the divergence would be different from the average. We then calculated the average value of ^(x) across each subset and divided by to obtain the relative age, A, compared with the autosomal average (Supplementary Table 6). We began by considering subsets of the genome consisting of the neighbourhoods of HC sites. The subsets SHC(d) were defined as all bases within a distance d of an HC site on an autosome (excluding the site itself, to obtain unbiased estimates of local genetic divergence). We expect that human and chimpanzee would be more closely related near HC sites. Analysis of the HCGOM shotgun data shows that relative age decreases with d (Fig. 2) and reaches a limit of A 0.862 0.009—that is, 86.2% of the average genetic divergence across the autosomes (Supplementary Note 4). Because the human–chimpanzee genome divergence time is thought to be about 7 Myr ago (refs 5, 10), this translates to a roughly 1-Myr reduction. The true difference must be even larger, because the average of (x) near HC sites must be greater than species. Figure 2: Near a region of HG, CG or HC clustering, (x) deviates strikingly and significantly from the genome average. a, For the HCGM shotgun data, the human–chimpanzee divergence is less than 88% of the genome average near an HC site (upper solid line) and at least 124% of the genome average near an HG or CG site (lower solid line). b, For the HCGOM shotgun data, the observed proportions are 86% and 134% (Supplementary Note 4). The true range is certainly larger than shown in either graph because HG and CG sites resulting from recurrent mutation (which are more frequent for HCGM than HCGOM data) dilute the estimated increase in human–chimpanzee divergence near these sites. Correcting for this by using a modelling analysis fitted to the HCGOM shotgun data (Supplementary Table 11) indicates that the divergence near true HG and CG clusters might be about 147% of the genome average. A dotted line corresponding to this extrapolated divergence is shown. (No line is shown for HC because it is very similar to the unextrapolated line.) Each data point is obtained by averaging all bases within a window geometrically centred on distance d (d/10 to d10). Error bars here and in Fig. 3 give 1 standard error. Second, we considered the neighbourhoods of HG or CG sites. The subsets SHG/CG(d) were defined as all bases within a distance d of either an HG or a CG site. We would expect that human and chimpanzee would be more distantly related near such sites. The relative age increases with d (Fig. 2) and reaches a limit of A 1.342 0.022 (Supplementary Note 4). Near two HG or CG sites the increase is even greater: A 1.45 0.07. This is primarily because two HG or CG sites close together are less likely to reflect recurrent mutations (comprising about 32% of HG and CG sites in the HCGOM data; Supplementary Table 5). When we use modelling to extrapolate the value of A near HG or CG sites in the absence of recurrent mutation, we infer that the true limit is even greater: A 1.47 (Methods). Large reduction in divergence time on chromosome X Third, we considered the divergence along individual chromosomes, and especially chromosome X. The relative divergences for the autosomes are all close to the average (Fig. 3), but divergence is reduced along nearly the entire length of chromosome X (Fig. 3). A slightly reduced age for chromosome X is in fact expected from population genetic theory; the population size of chromosome X should be three-quarters of that of the autosomes, and thus the coalescent time should be three-quarters as large. Calculations would predict A 0.918–0.943 (Supplementary Note 5), but the observed value is much younger: A 0.835 0.016 (Supplementary Table 6). a, Human–chimpanzee genetic divergence is lower on chromosome X than on every other chromosome (HCGOM shotgun data, correcting for recurrent mutation), and lower than the theoretical expectation of about 93% for a constant-sized, freely mixing ancestral population (Supplementary Note 5). By contrast, the gorilla chromosome X comparison shows no decrease beyond the expectation of about 95% from theory. We have included a category pooling all chromosomes of less than 100 Mb in size, because the smaller chromosomes do not have much data and thus have large standard errors. b, The low time divergence is seen along nearly the entire X chromosome. Black symbols show the ratio of human–chimpanzee to human–macaque divergence plotted in non-overlapping 5-Mb bins in the 15.1 Mb of HCOM alignment. For comparison we also show the ratio of human–chimpanzee to human–gorilla divergence (10-Mb bins) in 372 kb of HCGOM shotgun data (red, 1 standard error bars removed for clarity). c, The rate of HG and CG sites (HCGOM shotgun data, 10-Mb bins) is also greatly reduced along chromosome X, which is consistent with humans and chimpanzees being most closely related essentially everywhere along chromosome X. To confirm the low divergence of chromosome X, we performed the same analysis for human–gorilla divergence and found no discrepancy between the expected A 0.932–0.958 (Supplementary Note 5) and the observed A 0.977 0.028 (Supplementary Table 6). This excludes the possibility that the reduced divergence in the human–chimpanzee comparison reflects a slowdown of the mutation rate on chromosome X in the apes, because this would affect the gorilla comparison as well. As an independent line of evidence, we note that if human–chimpanzee divergence on chromosome X is recent, we would not expect segments of human–gorilla or chimpanzee–gorilla clustering (Fig. 1c, d). In fact, the rate of HG and CG sites is one-quarter of the autosomal rate. The rate is slightly lower than would be expected if the sites were due entirely to recurrent mutation; the 95% credible interval for the proportion of HG and CG sites not due to recurrent mutation is 0–15% of the autosomal rate (Supplementary Note 6). The data are consistent with a complete absence of regions where humans and gorillas, or chimpanzees and gorillas, are most closely related. The data point to an enormous decrease in genetic divergence for chromosome X in comparison with the autosomes. On the basis of A 0.835 0.016 for chromosome X (Supplementary Table 6) and calibrating to an estimate of human–chimpanzee genome divergence about 7 Myr ago2, 5, the average age difference between chromosome X and the autosomes must be about 1.2 Myr. The age difference between chromosome X and the autosomes in humans today is an order of magnitude smaller (Supplementary Note 7), again indicating an unusual history in the ancestral population at the time of speciation. We note that the reduced divergence of humans and chimpanzees on chromosome X also resolves a controversy about mutation rates in males versus females. Comparisons of genomes have shown a lower rate of sequence divergence on chromosome X than the autosomes for many species. On the basis of the hypothesis that this reflects a lower mutation rate in the female germline, it has been used to estimate the ratio of male:female mutation rates. Studies of human genome repeats11 and human–rat comparisons14 have indicated that 1.9–2.1, but comparisons of human and chimpanzee12, 15 have indicated that 6–7. The discrepancy can be resolved if the low human–chimpanzee divergence on chromosome X reflects a low time divergence. Correcting for this, we estimate that 1.9 (Supplementary Note 8), giving no evidence for an increase in on the primate lineage16. Implications for current models of human–chimp speciation The inference that human–chimpanzee genetic divergence varies over more than 4 Myr, and that genetic divergence is about 1.2 Myr less on chromosome X than the autosomes, raises two issues about human–chimpanzee speciation. First, these results place an upper bound on the age of human–chimpanzee speciation that poses conflict with some inferences from the fossil record. The Toumaï fossil (Sahelanthropus tchadensis), with its bipedalism and hominin dental features, is usually interpreted as being on the hominin line and setting a minimum date for human–chimpanzee speciation17, 18. The fossil was originally dated to 6–7 Myr ago (refs 17, 18), and a more recent study estimates 6.5–7.4 Myr ago (ref. 19). We compared inferences for the human–chimpanzee speciation date based on the Toumaï fossil with those that would be obtained by extrapolating from older species divergences. We first used an extreme upper bound of 20 Myr ago for human–orangutan genetic divergence20 (Supplementary Table 7). On the basis of the relative genetic divergence of human and chimpanzee (Supplementary Tables 7 and 8; Supplementary Note 9), we infer genome < 7.6 Myr ago and, from the bound species < 0.835genome, we can infer that species < 6.3 Myr ago. Using a more realistic estimate of human–orangutan genome divergence of less than 17 Myr ago, we obtain a younger bound of species < 5.4 Myr ago. The first bound is not compatible with the older range for Toumaï . The second bound is difficult to reconcile not only with interpretations of Toumaï but also with other fossils recognized as early hominins: Orrorin tugenensis at about 5.8 Myr ago (ref. 21) and Ardipithecus kadabba at about 5.6–5.8 Myr ago (ref. 22). We emphasize that these calibrations to the older fossil record are not likely to be compromised by molecular clock errors: a 'rate test' shows evidence of only slight lineage-specific changes in the mutation rate since the divergence of the great apes23 (Supplementary Tables 8 and 9). Similar bounds on human–chimpanzee speciation time are obtained by calibration to macaque fossil divergence (Supplementary Note 9). Second, the properties of chromosome X indicate an unusual evolutionary history around the time of human–chimpanzee ancestral speciation—proving that the structure of the population around the time of speciation was unlike that in any modern human or apes. In a freely mixing population under neutral drift, the ratio R of the genetic divergence on chromosome X and the autosomes should be about 0.75 (the effective population size of chromosome X relative to the autosomes). Such values are, in fact, observed in humans (R 0.59–0.87) and chimpanzees (R 0.56–0.76) and inferred for the population ancestral to human and gorilla (R 0.68–1.00) and chimpanzee and bonobo (R 0.75) (Supplementary Note 10). By contrast, the inferred value of R in the population ancestral to human and chimpanzee is 0–0.29 (Supplementary Table 10). We were not able to devise a demographic history consistent with such a low R, even with models of asymmetry between the sexes (Supplementary Note 11). However, strong selection across chromosome X could produce this effect. The apparent conflict with interpretations of the fossil record could be explained if Toumaï were somewhat younger than previously reported19, or if there was a problem with the molecular clock used for the calibrations to older fossil divergences. These factors, however, would not explain the more than 4 Myr spread of genetic divergence times across the genome, or the evidence for intense natural selection on chromosome X. We note that, on general grounds, we might expect to see greater evidence for natural selection on chromosome X than the autosomes, because recessive genetic variants are subject to selection in hemizygous males. However, we see no evidence for an unusual X chromosome divergence in the human–gorilla comparison. Possible hybridization in the human–chimp lineage We suggest a provocative explanation for multiple features of these data: that the hominin and chimpanzee lineages initially separated but then exchanged genes before finally separating less than 6.3 Myr ago (Fig. 1e). First, this could explain how Toumaï could have dates older than hominin speciation and yet still have hominin features17, 18, 19. Second, it could explain the wide range of divergence times (more than 4 Myr): at some loci human and chimpanzee lineages share ancestry around initial separation, whereas at others the genetic ancestry is more recent at the time of hybridization. Third, it could explain the low divergence of human and chimpanzee on chromosome X. An empirically observed pattern, documented in Drosophila, tsetse flies, mosquitoes, butterflies and guinea-pigs24, is that "the genes having the greatest effect on hybrid sterility and inviability are X-linked"24. The reasons for this 'second rule of speciation'24 are not fully understood25, 26, 27, although they are thought to be related to Haldane's rule about hybrid sterility affecting the heterogametic sex more than the homogametic sex28. A corollary—not previously suggested—is that if gene flow between two diverged populations occurs, chromosome X should be subject to strong and rapid selection to eliminate alleles, from one parental population or the other, that contribute to reduced fitness. The presence of multiple hybrid incompatibility loci could lead to selection across much or all of chromosome X, as in our data (Fig. 3). As a specific example, if human and chimpanzee ancestors initially speciated and then interbred, hybrid males might have been infertile, consistent with Haldane's rule. A viable population could then only have arisen if the fertile females mated back to one of the ancestral populations (for example, chimpanzee ancestors), producing fertile male hybrids when they transmitted X chromosomes derived almost entirely from that ancestral population. This could explain why humans and chimpanzees are most closely related throughout chromosome X. We note that in wild mice in the European Mus musculus/domesticus hybrid zone there has been reported to be a gradient of genetic variants on the autosomes, but a sharp geographic transition for chromosome X (ref. 29). This indicates that wild mouse hybrid populations might have difficulty carrying X chromosomes from multiple ancestral populations, which is consistent with what would be expected from our model and proposed corollary to the second rule of speciation. Speciation in animals is generally believed to occur by allopatry—that is, by the formation of an isolation barrier with no subsequent gene flow. When subsequent hybridization does occur, it is generally believed that the resulting population dies out30, 31. However, there are known examples of adapted hybrid populations in nature32, 33, 34, and hybridization could be advantageous, allowing nascent species to derive traits from several ancestral populations, combining them to adapt to new environments35. The failure to observe more instances of successful hybridizations in field studies so far30, 31 might simply be due to ascertainment bias—the fact that hybridizations occur too episodically to be observed practically. With comparative genomic methods, one can project backwards in time to make inferences about what happened at the time at which speciation occurred. Allopatric speciation without subsequent gene flow predicts that population genetic structure before and after speciation should be similar30, 31. By contrast, hybridization predicts a wide range of divergence times and different coalescence times in parts of the genome, such as chromosome X. By comparing the genomes of modern species, one could systematically test whether hybridization is a widespread process in evolution. We have shown that human and chimpanzee speciation was complex; furthermore, our model makes predictions that can be tested with larger data sets36. First, it predicts that evidence of natural selection should be seen not only on chromosome X but also at some autosomal loci. Such ancient selective sweeps might be detected as long regions with unusually low (or high) rates of human–chimpanzee divergence and HG and CG sites. (It might even be possible to identify autosomal genes under selection). Second, if a hybridization involving a single episode of gene flow occurred, it might result in a bimodal distribution of (x). Third, speciation involving hybridization could give rise to distinctive patterns of 'frozen linkage disequilibrium': differences in the lengths and distribution of HC, HG or CG clustering (Fig. 2). All these hypotheses can be tested once the gorilla genome is complete and aligned to the genomes of humans, chimpanzees and more distant primates. Methods DNA sequence data We sequenced 115, 152 fragments of DNA ('shotgun reads') from a western lowland gorilla and 2, 710 from a black-handed spider monkey (Ateles geoffryi) (these data are publicly available in the NCBI trace archive, http://www.ncbi.nlm.nih.gov/Traces). We combined our data with whole-genome shotgun sequence data from orangutan and macaque from the Washington University Genome Sequencing Center, the Baylor College of Medicine and the Venter Institute (http://www.ncbi.nlm.nih.gov/Traces/trace .cgi; we thank our colleagues for making these data publicly available) (Supplementary Table 1). For a supporting data set, we analysed finished contiguous sequence from bacterial artificial chromosome (BAC) sequencing of sections of chromosomes 7 and X, generated by the NIH Intramural Sequencing Center13.Genome alignments All shotgun reads were aligned to the NCBI Build 34 human genome assembly using the Arachne37 or BLASTZ38 programs (Supplementary Table 1). At loci with at least 100 base pairs of DNA aligned across all species of interest, we used the Multiple Alignment Program39 to obtain optimized local alignments (Supplementary Methods). We then applied four filters (Supplementary Table 2 and Supplementary Methods) to eliminate the following: first, alignments with an extremely high rate of intraspecific polymorphism; second, alignments with an extreme rate of reads from any one species; third, alignments with a very high rate of divergent sites from some species; or fourth, alignments that mapped to known segmental duplications in humans or chimpanzees40. Application of these filters minimized misalignment in our data but did not qualitatively change our main inferences (Supplementary Table 3). The Threaded Block Set Aligner41 was used for alignment of the BAC data. Aligned data sets are available online and at our website (http://genepath.med.harvard.edu/~reich). Identification of divergent sites for analysis For the shotgun data, we used divergent sites only if they had a sequencing quality score of at least 30 and 5 bases on each side with quality 25 or more. We additionally required sites to be single nucleotide substitutions, to show exactly two alternate alleles across the species, to be outside hypermutable CpG dinucleotides, to be at least three base pairs from a repeat identified by Tandem Repeat Finder42, and to be flanked on either side by at least one base of perfect alignment across species. When multiple reads were available, we used data from the read with the highest sequence quality. Application of these filters did not qualitatively affect the main conclusions from our analysis (Supplementary Table 3). The filtered data sets are available online and at our laboratory website (genepath.med.harvard.edu/~reich). Genetic divergence estimates The main measurement in this paper is genetic divergence between two species. To calculate this over a stretch of sequence, we always counted the number of differences per base pair between the two species and normalized by the difference between human and macaque (or another outgroup) over the same stretch. This corrects for variability in mutation rate from locus to locus. In particular, it corrects for a high local mutation rate. As an example for the HCGOM data, the normalized divergence is proportional to (x) = (nH + nHG + nC + nCG)/(nH + nHC + nHG + nHCG + nM). Using human–macaque divergence for normalization is slightly conservative for X–autosome comparisons. Like all species pairs, humans and macaques are slightly less time-diverged on chromosome X; if we could correct for this, human–macaque divergence on chromosome X could be up to a few per cent less than shown in Supplementary Table 6. Acknowledgments We thank B. Bodamer, J. Caswell, M. Clamp, J. Coyne, J. Cuff, E. Green, G. McDonald, J. Mullikin, H. A. Orr, D. Page, D. Pilbeam, N. Stange-Thomann and M. Zody for discussions, comments and assistance with stages of this study, and E. Green, R. Gibbs and R. Wilson for producing and making publicly available data from large-scale sequencing projects (contiguous data from chromosomes 7 and X, and the orangutan and macaque shotgun data). N.P. was supported by a career transition award from the National Institutes of Health. E.S.L. was supported in part by funds from the National Human Genome Research Institute and the Broad Institute of Harvard and the Massachusetts Institute of Technology. D.R. was supported in part by a Burroughs–Wellcome Career Development Award in the Biomedical Sciences.Author Contributions The authors all played significant roles in the conception, execution, interpretation and presentation of the study. Competing interests statement: The authors declared no competing interests. Supplementary information accompanies this paper. |