Time for mankind to take a shot at designed intelligence?

Can a Dog Be Bred to Be as Smart as a Human? 

Denyse O'Leary 

Within one hundred generations or roughly 600 years? That’s the project Payton Pearson, an electrical engineer who gives his affiliation as the Air Force Institute of Technology in Dayton, Ohio, has set himself: 

Artificial selection is a well-known phenomenon of selecting for certain physiological characteristics of various species of plants and animals, and it is something that human beings have been doing for thousands of years. A perfect example of this is the union and development of dogs under human stewardship since the beginning of the agricultural era of society. In that time, approximately 6,000 years [1], dogs have been artificially selected in such a way as to produce thousands of different breeds. From the stout Dachshund, a dog breed produced for the purpose of hunting den-dwelling animals, to the highly intelligent Border Collies who were bred to help sheep herders herd their flocks; the many different canine breeds have served humanity in a multitude of capacities for many generations. In this paper, using the concept of artificial selection,[1] it is determined with mathematical and statistical evidence how humans could artificially select for canine intelligence to such a degree as to produce canines with human levels of intelligence within a relatively short amount of time — 600 years. 

PAYTON PEARSON, “ARTIFICIALLY SELECTING FOR INTELLIGENCE IN DOGS TO PRODUCE HUMAN-LEVEL IQ WITHIN 100 GENERATIONS” AT FIGSHARE (SEPTEMBER 24, 2022) 

 Praise for the Border Collie 

He points to the Border Collie, considered to be among the most intelligent of breeds: 

If one were to roughly compare the extent to which Border Collies have had their intelligence improved, it equates to a dog with the intelligence of a three-year-old child, or about 40% more intelligent than an average dog—which is to say that an average Border Collie can be considered a genius of the dog world; this means they would equate to a human with an IQ of 140 or higher on average.

PAYTON PEARSON, “ARTIFICIALLY SELECTING FOR INTELLIGENCE IN DOGS TO PRODUCE HUMAN-LEVEL IQ WITHIN 100 GENERATIONS” AT FIGSHARE (SEPTEMBER 24, 2022) 

He rates the Border Collie as 35 in human IQ terms (the human average being 100) and sets out a program for selective breeding of smarter dogs, which readers may peruse at leisure. 

Up Front About His Goals 

Pearson writes:

The main barrier here, above all else, is the proper advertising and advocation for such cultural moves. Being that these ideas are seemingly taboo in so many cultural circumstances, the zeitgeist needs to be shifted as such, which is one of the primary purposes of this paper. 

PAYTON PEARSON, “ARTIFICIALLY SELECTING FOR INTELLIGENCE IN DOGS TO PRODUCE HUMAN-LEVEL IQ WITHIN 100 GENERATIONS” AT FIGSHARE (SEPTEMBER 24, 2022) 

As to the limitations of his methods: 

Converting the IQ from one species to another is difficult, and very much an imperfect science. 

PAYTON PEARSON, “ARTIFICIALLY SELECTING FOR INTELLIGENCE IN DOGS TO PRODUCE HUMAN-LEVEL IQ WITHIN 100 GENERATIONS” AT FIGSHARE (SEPTEMBER 24, 2022) 

We’ll come back to that in a minute. Also, 

If one looks at the Flynn Effect as measured in human populations [13], it has been empirically determined that average human IQ has been increasing by three points per decade for the last few hundred years, though it does appear to be slowing in current times, and with no well-understood reason. It could be for any number of possible exterior factors. Not only could there be coevolving factors slowing down the process, but it could also be the very exponential nature of IQ, and how its measured increase necessarily slows as the raw values increase. This would mean, perhaps, that IQ increases could be a measure of the efficiency of the cerebral cortex and how well-folded it is within its volume, rather than an increase in cranial volume itself, but in-depth discussion on anatomy is beyond the scope of this paper. 

PAYTON PEARSON, “ARTIFICIALLY SELECTING FOR INTELLIGENCE IN DOGS TO PRODUCE HUMAN-LEVEL IQ WITHIN 100 GENERATIONS” AT FIGSHARE (SEPTEMBER 24, 2022) 

Well, it’s too bad if anatomy is beyond the scope of the paper because it is a critical factor here. Would a dog brain support a human-like neurological capability? It would require major restructuring of the brain in an animal that is a canid, not a primate. If we can’t do it with chimpanzees, why should we think we can do it with dogs? 

The Flynn Effect 

About the Flynn effect in humans: It’s real but it is most likely the result of gradually improving standards of living, medical care, and schooling. People who are healthy, well-fed, and accustomed to learning might be expected to perform better on intelligence tests than their less fortunate ancestors. But, as Pearson notes, the process slows down because, like everything else — including dog intelligence — it can only go so far before it reaches a ceiling. 

Cultural advantages like a good science education might make people competent science teachers or researchers but it doesn’t turn them into Albert Einstein (1879–1955), Kurt Gödel (1906–1978), or Marilyn vos Savant. For the same reasons, Pearson may be able to breed a smarter dog but not a dog that thinks like a human.

In any event, human consciousness is famously the “Hard Problem of Consciousness.” We don’t even know what it is, exactly.

Thus, with dogs, we don’t know what we are aiming at and it is not clear in that case how multiplying generations will get us anywhere. 

Dogs Are Already Smart, as Animals Go 

All that said, the American Kennel Club advises that dogs are smarter than we think without any major changes anyway: “They can also understand more than 150 words and intentionally deceive other dogs and people to get treats, according to psychologist and leading canine researcher Stanley Coren, PhD, of the University of British Columbia.” (2009) Coren, who writes the “Canine Corner” column at Psychology Today, agrees with Pearson that Border Collies are the smartest dogs: “Border collies are number one; poodles are second, followed by German shepherds. Fourth on the list is golden retrievers; fifth, dobermans; sixth, Shetland sheepdogs; and finally, Labrador retrievers,” said Coren.

Here’s an attempted ranking of dog breeds by intelligence and other traits. 

Charles Fawole tells us at Pet Calculator (April 14, 2022), that there are three types of dog intelligence we can measure: Instinctive intelligence (for hunting, guarding, or herding, for example), which tends to be inherited based on the breed; adaptive intelligence, which is the dog’s individual ability to learn things on its own; and working/obedience intelligence, which is what a dog can do with instruction from humans. According to Fawole, the dog that takes the longest to learn commands is the Afghan hound. 

Incidentally, according to Jean Marie Bauhaus at Hills Bros, a pet food company site, large dogs are smarter than small ones: The suggested explanation is that the small dogs often have oddly shaped heads, which may interfere with brain development: “Meanwhile, mesocephalic dogs — those with average-shaped heads, like Labrador retrievers — tend to lack such specialization, which, according to researchers, may give them more cognitive flexibility that makes them better at learning new tasks.” (June 12, 2019)

 Reason and Moral Choice 

But comparisons with humans, as in “When comparing Border Collies to humans studies have shown that they show intelligence close to a 3-year-old child,” are off base. The child is becoming a creature with reason and moral choice and the dog isn’t. Keeping that in mind will help us navigate a complex living world more realistically. 

You may also wish to read: A science writer makes the case for plants as conscious beings. Annaka Harris, neuroscience and physics writer, starts by casting doubt on human consciousness. But if it’s true that we can’t trust the reasoning skills of our evolved brains to arrive at a correct answer, we are in no position to evaluate Harris’s own thesis.

The fossil record fossil recording some more.

Fossil Friday: The Giant Armadillo Glyptodon and the Abrupt Origin of Xenarthrans

Günter Bechly 

This Fossil Friday we continue our series on the origins of the various modern placental mammal orders. Our subject today is the order Xenarthra, which is endemic to the New World and divided into two suborders, Cingulata (armadillos and their extinct relatives, pampatheres and glyptodonts) and Pilosa (sloths and anteaters). Both suborders include examples of the extinct Pleistocene megafauna of America. The suborder Pilosa includes the giant ground sloths like the genus Megatherium, and the suborder Cingulata includes the car-sized giant armadillos of the genus Glyptodon. A complete skeleton of Glyptodon asper from the Pleistocene of Argentina is featured at the top of this article. This specimen is on display at the Natural History Museum of Vienna. With an age of only 1.5 million years it is very far from being the oldest representative of this order. A recent review article by De Iuliis (2018) commented that: 

These three clades and the differences between them reflect deep histories that reach back at least to the Paleocene and their fossil representation (of cingulates and sloths, but not vermilinguans) is exceptionally rich and diverse. 

So, let’s have a closer look at this rich fossil record of xenarthrans. Gaudin & Croft (2015: fig. 2) provided an excellent chart of the stratigraphic distribution of the xenarthran fossil record with up-to-date chronology of the South American Land Mammal Ages (SALMA). They concluded that ”no unambiguous Paleocene fossil xenarthran remains are known.” So what are the oldest fossils of this order known to science? 

Oldest Fossils of the Order 

The more primitive suborder Cingulata is indeed well represented in the Paleogene fossil record from South America (Sedor et al. 2022). The earliest known cingulate and arguably oldest fossil representative of Xenarthra is Riostegotherium yanei described by Oliveira & Bergqvist (1998) from the Itaboraí Basin in the Rio de Janeiro State of Brazil. This material was already mentioned by Scillato-Yané (1976) under the incorrect name Prostegotherium. There is also at least one other undescribed armadillo-like species from this locality and age (Cifelli 1983, Bergqvist et al. 2004, 2019). A thorough revision of the material by Bergqvist et al. (2004) confirmed the identification of Riostegotherium as oldest xenarthran remains (also see Bergqvist et al. 2019). 

Interestingly, Bergqvist et al. (2004) also observed that “shared similarities to palaeanodonts add support to the proposal that Palaeanodonta may have been ancestral to, or is the sister-taxon of Xenarthra”. Why is this interesting? Because the extinct Palaeanodonta a few years later turned out to be stem pangolins and both turned out to not be related to Xenarthra at all (e.g., Gaudin et al. 2009). This shows how the alleged connection between morphological similarity and phylogenetic relationship is rather a Darwinist myth that is upheld as convenient tool to promote the theory.

Anyway, there are also some problems with the dating of Riostegotherium: The deposits where it was found belong to the Itaboraian SALMA, which was was originally dated by Marshall (1985) to the middle Paleocene (63.6-61 mya), but this was indirectly inferred in the absence of radioisotopic or magnetostratigraphic datings. Bond et al. (1995) instead suggested a Paleocene/Eocene age of 60-57 mya, which was also accepted by Bergqvist et al. (2004, 2019). Gelfo et al. (2009) provisionally re-dated the Itaboraian SALMA to an Early Eocene age of 55.5-53.5 mya (Lower Ypresian). The seminal work of Woodburne et al. (2014) moved the dating even a bit younger to the Middle Ypresian about 53-50 mya, contemporaneous with the Early Eocene Climatic Optimum (EECO). Similarly, Gaudin & Croft (2015) suggested an age of 54-52 mya in their stratigraphic chart. Nevertheless, Bergqvist et al. (2019) mentioned that “study in progress by the senior author challenges this latter 

interpretation, suggesting that at least part of the fissures was formed and filled during the late Paleocene.” Therefore they just gave a vague Paleocene/Eocene age. Apparently the study in progress is not yet published to this day. Therefore, it cannot yet be excluded that Riostegotherium might be of Late Paleocene age after all.

Prior to the discovery of Riostegotherium, the oldest xenarthrans were astegotheriine armadillos from the Barrancan subage of the Casamayoran SALMA of Argentina, which was believed to be about 7 million years younger (Scillato-Yané 1976). According to Kay et al. (1999) “the Casamayoran SALMA was at least 18–20 m.y. younger than previously assumed”. Therefore, Dunn et al. (2013) and Ciancio et al. (2016) dated the Barrancan to a Middle Eocene (Lutetian-Bartonian) age of 41.7-39.0 mya, which was basically confirmed by Woodburne et al. (2014) with an estimate of 42-38.5 mya (also see Gaudin & Croft 2015).

Remains of the astegotheriine armadillos Stegosimpsonia sp. and Astegotherium dichotomus were recovered from Cañadón Vaca in Patagonia. The respective layers date to the older Vacan subage (as opposed to the younger Barrancan subage) of the Casamayoran SALMA from the early Middle Eocene about 46-44 mya (Cifelli 1985, Gaudin & Croft 2015, Ciancio et al. 2018, 2019).

The extinct armadillo taxa Peltephilidae and Astegotheriini are also known from the Riochican SALMA of Patagonia (Simpson 1948, McKenna & Bell 1997, Oliveira & Bergqvist 1998, Gelfo et al. 2010). The Riochican SALMA is interspersed between the older Itaboraian SALMA and the younger Casamayoran SALMA. It had been dated to a Late Paleocene age of 57-55.5 mya by Bond et al. (1995), but was later re-dated to be only about 49-48.5 mya (Ypresian, Early Eocene) according to Woodburne et al. (2014), and to 51-49.5 by Gaudin & Croft (2015).Armadillo remains of Pucatherium parvum and Noatherium emilioi have recently been described from the lower Lumbrera Formation in Northwestern Argentina (Fernicola et al. 2021), which the latter authors re-dated to the Early Eocene Climatic Optimum (Ypresian). This is older than previously believed, so that these armadillos could be contemporaneous with Riostegotherium if the re-dating of the Itaburaian by Woodburne et al. (2014) is correct (see above). 

Considerable Scientific Controversy 

Another recent discovery is a xenarthran metacarpal bone from the Late Eocene of Seymour Island in Antarctica (Davis et al. 2020). It was found in the Cucullaea I Allomember (TELM4) of the La Meseta Formation. The age of this formation is a matter of considerable scientific controversy, which I will discuss in great detail in a forthcoming technical paper on the waiting time problem in the origin of whales (Bechly et al. in prep.). Here it must be sufficient to note that an Early Eocene (Ypresian) age of 52-48 mya seems most likely and represents the consensus of most experts based on different lines of evidence (strontium isotopes, magnetostratigraphy, eustatic lowstands, dinoflagellate and mammalian biostratigraphy), while some dissenters are considering a younger Middle Eocene (Bartonian) age of about 40 mya mainly based on some dinoflagellate cysts.

The earliest fossil record for the suborder Pilosa is Pseudoglyptodon from the Priabonian-Bartonian (37.8 mya) of Argentina (Gaudin & Croft 2015). Within Pilosa there are two major subgroups: sloths (Folivora) and anteaters (Vermilingua). Ameghino (1895) described an astragalus of Proplatyarthrus longipes from the early Late Eocene (Mustersan SALMA), which could be the oldest folivoran, but unfortunately this specimen seems to be lost according to Pujos et al. (2021). The oldest other alleged sloth remains were from the Middle Eocene of Antarctica (Vizcaíano & Scillato-Yané 1995), but were disputed and considered to be an undetermined mammal by MacPhee & Reguero (2010). Pseudoglyptodon sallaensiswas described by Engelmann (1987), based on a mandible fragment from the Oligocene Salla Beds (Deseadan SALMA) of Bolivia, which also yielded other armadillo fossils (Billet et al. 2011). As already mentioned 

before, Pseudoglyptodon is commonly considered to be the oldest folivoran, but its possibly not a sloth in the strict sense (Pujos & De Iuliis 2007). Due to its peculiar characteristics and fragmentary preservation it remains enigmatic (Pujos & De Iuliis 2007, Pujos et al. 2021). Engelmann (1987) gave an age of 25-28 mya, which agrees with modern datings of the Deseadan SALMA to 29.4-24.2 mya (Dunn et al. 2013). McKenna et al. (2006) described new material of Pseudoglyptodon from the Late Eocene-Early Oligocene (Tinguirirican SALMA, 33.6-31.3 mya, Dunn et al. 2013) of Chile and the Late Eocene of Cerro Blanco in Argentina (Mustersan SALMA), and recognized this genus as the sister group of all other sloths. Nevertheless, Varela et al. (2019) commented: “However, as Pseudoglyptodon presents many convergences with cingulates, we cannot rule out its position to be an artifact of our limited knowledge of this taxon.” The Mustersan SALMA was traditionally considered to be of Middle Eocene age, then considered to be postdating 35-36 mya (Kay et al. 1999); but ultimately recognized as early Late Eocene with an age estimates of 38.2-38 mya (Dunn et al. 2013), and 37.9-36.5 mya (Woodburne et al. 2014, Gaudin & Croft 2015). Consequently, the Argentine material 

of Pseudoglyptodon indeed represents the oldest known fossil pilosans and likely the oldest sloths as well.

Some other very early fossil remains that are unequivocally referable to sloths are a Megalonychidae from the Early Oligocene (ca. 35 mya) of Puerto Rico (MacPhee & Iturralde Vinent 1995) and Patagonia (Carlini & Scillato-Yané_2004).

What About Anteaters? 

The oldest fossil record of anteaters is an undescribed vermilinguan from the Early Miocene Colhuehuapian SALMA (20 mya) of Patagonia (Carlini et al. 1992, Gaudin & Branham 1998, Gaudin & Croft 2015). This implies a 30 million year ghost lineage of undocumented existence (Delsuc et al. 2001). The oldest described vermilinguan is Protamandua rothi from the Early to Middle Miocene Santacruzian of Patagonia (Ameghino 1904, Hirschfeld 1976, Patterson et al. 1992, Gaudin & Branham 1998). According to Gaudin & Croft (2015) it is about 18 million years old, but PaleoDB gives a range of 17.5-11.608 mya. We see that in most cases even the most modern radiometric methods provided quite different and fuzzy datings of the fossil outcrops. This does not mean that all the datings are completely wrong, but it shows that we should not place too much confidence in the current consensus concerning precise dates and possible evolutionary scenarios that are based upon them.

This also holds for so-called molecular clock datings, which place the origin of armadillos around the K/Pg boundary (Delsuc et al. 2001, 2004, Presslee et al. 2019). For Darwinists this would necessarily require an even earlier origin for the order Xernarthra in the Late Cretaceous. Too bad, this is of course strongly contradicted by the fossil record and the total absence of any Cretaceous crown group placental mammals. Should we dare to consider the possibility that something is wrong with the Darwinist assumptions? Heaven forbid!

An Abrupt Appearance 

Anyway, we can conclude from the discussion above that the order Xenarthra appeared abruptly in the Lower Eocene about 53 million years ago, only 3 million years after the oldest fossils of our own order Primates, which we discussed last week. Carlini & Scillato-Yané (2004) made a revealing admission: 

The absence of most major groups of Xenarthra in South American mammal-bearing sediments of Late Cretaceous and Early Paleocene age is the greatest enigma in the study of this Superorder. 

I totally agree and can only note that this of course only emphasizes the abruptness of their appearance in the fossil record.

Xenarthra is one of the four major clades of placental mammals recognized in modern phylogenetic systematics. The three others are Afrotheria, Euarchontoglires, and Laurasiatheria. Next Fossil Friday we will look at a member of the African mammal clade Afrotheria, i.e., the aardvark order Tubulidentata and its fossil relatives.

P.S.: I have already discussed the misidentified alleged European anteater Eurotamandua from the Eocene Messel Pit in a previous Fossil Friday article (Bechly 2022).