Friday 30 November 2018

Statistical pitfalls of personalized medicine (2018)

An excellent paper in Nature by Stephen Senn. The article begins :

Personalized medicine aims to match individuals with the therapy that is best suited to them and their condition. Advocates proclaim the potential of this approach to improve treatment outcomes by pointing to statistics about how most drugs — for conditions ranging from arthritis to heartburn — do not work for most people. That might or might not be true, but the statistics are being misinterpreted. There is no reason to think that a drug that shows itself to be marginally effective in a general population is simply in want of an appropriate subpopulation in which it will perform spectacularly.

The reasoning follows a familiar, flawed pattern. If more people receiving a drug improve compared with those who are given a placebo, then the subset of individuals who improved is believed to be somehow special. The problem is that the distinction between these ‘responders’ and ‘non-responders’ can be arbitrary and illusory.

Much effort then goes into the effort to uncover a trait to explain this differential response, without assessing whether or not such a differential exists. I think that this is one of many reasons why a large proportion of biomarkers thought to distinguish patient subgroups fall flat. Researchers need to be much more careful.

More HERE.  

Thursday 29 November 2018

Intangible Cultural Heritage of Humanity (2018)

Reggae music has been recognised by UNESCO as part of the Intangible Cultural Heritage of Humanity.

It's about time they did. I have loved reggae music since June 1978 when I first heard King Tubbys Meets Rockers Uptown by the wonderfully named Augustus Pablo. I first heard this legendary track on an Island Records limited edition 12" single. The track is   2' 30" of pure thunder and lightening, not just pure Jamaican reggae, but one of the most stunning Jamaican dub reggae tracks of all time.


The Jamaican electronic and sound engineer Osbourne Ruddock (1941-1989), was born and raised on High Holborn Street, which is close to the harbour in Kingston Jamaica. As a teenager, Ruddock began studying electronics through American correspondence courses and at the College of Arts, Science and Technology in Kingston. After leaving college he set up a radio repair shop in Kingston, and from his repair work he began to take an interest in repairing and making electronic equipment and amplifiers for local recording studios and sound systems. Ruddock was technically gifted as an electronics engineer: in 1961 he built his own radio transmitter, from which he briefly broadcast a pirate radio station. 

Throughout the 1960s, Ruddock built a reputation as an electronics engineer, a sound-system operator, and a remixer of vocal reggae tracks. By 1971, he had converted the front room of his house on Dromilly Avenue, Kingston, into a recording and remix studio, and had begun releasing re-worked versions of vocal reggae tracks under his stage-name of King Tubby

Within this modest physical space, King Tubby taught himself how to play his studio as a musical instrument. He explored how his analogue mixing deck, sound effects and tape recorder could be used to strip away the original structure of a reggae vocal track, leaving fragments of the original vocals, keyboards or horns, and emphasising the percussive drive of the drums and the thunder of the original bass line. This process stripped back the layers of a song’s body, to leave exposed both the skeleton and viscera of the track.  Through a painstaking process of musical prosection, King Tubby helped to create dub reggae. Much of what is now known as remix culture in modern music can be traced back to the sonic inventiveness of Ruddock: playing with tape delay, high and low pass filters, echo, reverb, phasing, and modulation, in his home-made studio in Dromilly Avenue, Kingston.

Some of Ruddock's most distinctive albums are collaborations with the producer and multi-instrumentalist Augustus Pablo (1953-1999). Pablo was born Horace Swaby, in the middle-class Havendale district of Kingston, Jamaica. He attended Kingston College as a teenager, where he learned to play piano. By the late 1960's, Swaby had become a committed Rastafarian and had begun working as a session musician for the Randy's Studio house band. In 1969 he began playing melodica - a small, cheap, mouth blown keyboard that was made for teaching children basic keyboard skills. He released his first instrumental tracks in the early 1970s under the pseudonym Augustus Pablo, and from then until his early death in 1999, he produced hundreds of vocalists, and released a string of haunting reggae instrumental albums featuring his trademark Far East sound, based on minor key melodica motifs.

The collaboration between King Tubby and Augustus Pablo is exemplified in the album King Tubbys meets Rockers Uptown, which was first released on the Yard label in Kingston in 1976, though individual tracks had been released slightly earlier as singles. This is a foundational record in the development of Jamaican music. It is considered by many experts to be one of the greatest of all dub reggae albums. By construction, each of the 11 short tracks on the album is a radical re-working of an original vocal track. The combined talents of King Tubby and Augustus Pablo mean that not only were the original vocal tracks stripped back, but new rhythms were created by King Tubby through echo and reverb, and subtle shifts in mood were created by Pablo. This music has had a profound influence on modern music, as much for the remix approach it took to re-working an existing musical track in an inventive way than through musical similarity.

Tubby built his studio using mixing desks that were discarded and considered obsolete by Federal recording studio. He modified these decks by adding unique electronic circuitry of his own design, and then he and his collaborators ET and Lee Perry, explored the outer edges of what this relatively modest technology was able to do. Pablo did the same with the low-tech melodica keyboard. Tubby did not record musicians. He took pre-recorded musical tracks and added vocals and some over dubbed music - such as the melodica motifs of Pablo. He pushed to the outer limits the capability of the recording technology he had at his disposal. 

The writer Michael Veal has written a detailed, informed and serious history of the different threads of Jamaican culture and music that led to dub reggae. In this history, he notes that the way basic studio recording technology was used by people like King Tubby led to a fundamental change in Jamaican music: 

At this stage… recording technology was beginning to imply for reggae what improvisation had already implied for jazz: the notion that a ‘composition’ must now be understood as a composite of its endlessly multiplying, mutating, and potentially infinite elaborations over time.

Scroll Back Media (2018)

One of the prime qualities of books is the ability to repeat themselves for the reader, at the reader’s request, as many times as wanted.

Monday 26 November 2018

Hammers (1935)

Assorted modern hammers, from Art and Industry: The principles of Industrial Design by Herbert Read. 


Saturday 24 November 2018

Peacock feather (1861)

Anna Atkins and Anne Dixon: Peacock, from a presentation album to Henry Dixon, 1861.
From a good write up in the New York Review of Books on the pioneering cyanotypes of Anna Atkins. (HERE).


From Robert MacFarlane's word of the day series on his Twitter feed. One of my favourite northern words, Nesh

Data Fallacies to Avoid (2018)

An excellent poster HERE from Geckoboard. 

Thursday 22 November 2018

System Design, Modelling and Simulation (AD 150 - AD 2014)

System Design, Modelling and Simulation, edited by Claudius Ptolemy is  HERE. The Preface begins:

My last written work was published nearly 1,900 years ago. I am pleased to come out of retirement to give voice to a project that I’m proud to have named after me, the Ptolemy Project. Like much of my prior work in astronomy and geography, this project deals with complex systems. Like most of my prior writings, this text compiles the thinking and contributions of many people.

The motions of the planets, the sun, the earth, and the moon, which I studied in my work The Almagest, are concurrent interacting processes. They are deterministic, not subject to the whims of the gods. More accurately, the models that I developed, as well as those of many of my successors, deliberately ignore any effects that the gods might capriciously impose. These models focus instead on precisely matching observed behavior, and more importantly, on predicting behavior. The Ptolemy Project similarly studies concurrent processes and focuses on deterministic models.

Ideally, an intellectual quest moves human knowledge from superstition and unfounded beliefs to logic and measurement. What we now call “science,” particularly in the study of natural systems, is deeply rooted in the scientific method, where we form hypotheses, design experiments, and draw conclusions about the hypotheses based on the experiments. To be able to make measurements, of course, the artifact or process being measured must exist in some form. In my earlier studies, this was not an issue, since the sun, earth, moon, and planets already existed. Engineering disciplines, which focus on human-constructed artifacts and processes, however, study systems that do not yet exist. Nevertheless, the scientific method can and is applied in engineering design. Engineers construct simulations and prototypes of systems, formulate hypotheses, and perform experiments to test those hypotheses.

Because of the focus on artifacts and processes that do not yet exist, engineering design should not be based solely on the scientific method. The goal of experiments is to improve our understanding of the artifact or process being designed. But we have to create the artifact or process before we can perform experiments. Being forced to create something before we understand it dooms our design to roots in superstition and unfounded beliefs.

An important part of a science, quite complementary to the scientific method, is the construction of models. Models are abstractions of the physical reality, and the ability of a model to lend insight and predict behavior may form the centerpiece of a hypothesis that is to be validated (or invalidated) by experiment. The construction of models is itself more an engineering discipline than a science. It is not, fundamentally, the study of a system that preexists in nature; it is instead the human-driven construction of an artifact that did not previously exist. A model itself must be engineered.

Good models can even reduce the need for measurement, and therefore reduce the dependence on the scientific method. Once we have a model of the motions of the planets, for example, that we know accurately predicts their positions, there is less need to measure their positions. The role of measurements changes from determining the positions of the planets to improving the models of their motions and to detecting capricious actions of the gods (something that engineers call “fault detection”)...

In short, in engineering, as opposed to science, models play a role in the design of the systems being modelled. As with science, models can be improved, but unlike science, so can the systems being modelled.

Every time I drink a glass of water, I get an image of my father drinking a glass of milk, looking at me with smiling eyes. (2018)

A piece by Sam Anderson in the New York Times Magazine 


Sunday 18 November 2018

History is the long process of outsourcing human ability... (2011)

Immediately prior to the Jeopardy match that the IBM Watson computer won, the American novelist Richard Powers ruminated on AI (HERE). 

An image of the area striata - the visual region of the human brain from Eye & Brain by Richard Gregory (HERE)

Saturday 17 November 2018

The Kilogram’s Long, Slow Climb to Harmony (2018)

A superb piece in the New Yorker (HERE) on the recent redefinition of the metre in terms of Planck's constant. It is a superb illustration of the fundamental question asked in measurement: Compared to What?

Thursday 15 November 2018

Shoot, Big O! (2018)

The incomparable John McPhee has a new book out (HERE). It includes an album quilt of very short pieces that have a jewel like quality. Below a piece on a basketball player:

Oscar Robertson, of basketball’s Cincinnati Royals, entered the hall and was immediately surrounded by at least 2,000 people, more than half of them adults. To get near him, they climbed over booths, broke down barricades, and temporarily paralyzed most of the exhibits in the show. National Shoes had engaged Robertson to make an appearance and sign autographs. Soon, Robertson was standing in a small “basketball court”—ten feet wide by twenty feet long—between a pair of backboards made of thin composition board and equipped with attached hoops of the sort that are sold in dime stores. The crowd seemed to surge like a throng in Saint Peter’s Square. A little boy, perhaps ten years old, stood beside Robertson, and Robertson handed him a basketball. The boy took a shot, and missed. Robertson retrieved the ball and handed it to him again. The boy shot again, and missed. Robertson leaned down and talked to the boy. Not just a word or two. He spoke into the boy’s ear for half a minute. The boy shot again. Swish. Robertson himself seemed reluctant to try a shot. The baskets were terrible, and—even if they had not been—a basketball player makes only about half his shots anyway. A few misses, and this crowd really would not have understood.

Moreover, Robertson was wearing an ordinary business suit, so his movements would be restricted. He signed a few autographs. “Shoot, Big O!” someone called out. Others took up the cry. “Shoot, Big O!” Robertson turned aside, and signed another autograph. “Shoot, Big O!” Robertson studied one of the baskets. This might have been a mistake, because there was no retreating now. Once a basketball player, with a ball in his hand, looks up at a basket, almost nothing can make him resist the temptation to take a shot. Robertson stepped back to a point about 17 feet from the basket and lifted the ball high, and a long set shot rolled off his fingers and began to arc toward the basket with a slow backspin. The crowd was suddenly quiet. Everybody watched the ball except Robertson, whose eyes never left the basket until the ball had dropped in. He shot again. Swish. Again. Swish. Five, six, seven in a row. There was no one else in the Coliseum now. Robertson—making set shots, jump shots, even long, graceful hook shots—had retreated from the crowd into the refuge of his talent.

Tuesday 13 November 2018

Adequacy is for archangels (2018)

A fascinating interview with the writer Tim Robinson HERE

One of his answers below: 

‘Walking’ has been competently anatomised by several cultural theoreticians recently. Leaving those analyses aside, the practice has been essential to my method. Once a wealthy friend with a big car offered to help me in my explorations of Connemara. Since I wanted to revisit a few remote glens I accepted, and we roared off. Then, ‘I must call in at that cottage,’ I said, and we squealed to a stop. I knocked at the door, but apart from a twitching curtain there was no response – whereas if I had sweated up the hill, fallen off my old bike at the gate, asked for a bucket of water to mend a puncture, etc., all the lore of the valley would have been forthcoming over tea in the kitchen. But even bicycling is inferior to walking in this context. To appear out of the thickets behind an Aran cottage, or scramble down from the bare moon-mountains of the Burren into a farmyard, is, I find, a disarming  approach, introducing me as obviously unofficial and dying for a cup of tea.

Friday 9 November 2018

Composition (No. 1) Gray-Red (1935)

The Art Institute of Chicago has just released more than 50,000 high resolution images of their superb collection as Public Domain images. They can be used for anything, by anyone, without limit. 

Above is a close up of Composition (No. 1) Gray-Red by Piet Mondrian. The full size image is HERE.

Thursday 8 November 2018

Prokaryotes: The unseen majority (1998)

Q:How many individual organisms exist on Earth?
A: About 10,000,000,000,000,000,000,000,000,000,000.

A superb paper HERE on this topic.

Prokaryotes: The unseen majority

William B. Whitman*†, David C. Coleman‡, and William J. Wiebe§

Departments of *Microbiology, ‡Ecology, and §Marine Sciences, University of Georgia, Athens GA 30602


The number of prokaryotes and the total amount of their cellular carbon on earth are estimated to be 4–6 x 10^30 cells and 350–550 Pg of C (1 Pg = 10^15 g), respectively. Thus, the total amount of prokaryotic carbon is 60–100% of the estimated total carbon in plants, and inclusion of prokaryotic carbon in global models will almost double estimates of the amount of carbon stored in living organisms. In addition, the earth’s prokaryotes contain 85–130 Pg of N and 9–14 Pg of P, or about 10-fold more of these nutrients than do plants, and represent the largest pool of these nutrients in living organisms. 

A Visual History of Computing: 1945 - 1979


Computers, as artworks and artefacts. HERE.

Tuesday 6 November 2018

Unskilled and Unaware of It (1999)

HERE is a classic paper, introducing what is now known as the the Dunning-Kruger effect, after psychologists David Dunning and Justin Kruger.

Unskilled and unaware of it: how difficulties in recognizing one's own incompetence lead to inflated self-assessments.


People tend to hold overly favorable views of their abilities in many social and intellectual domains. The authors suggest that this overestimation occurs, in part, because people who are unskilled in these domains suffer a dual burden: Not only do these people reach erroneous conclusions and make unfortunate choices, but their incompetence robs them of the metacognitive ability to realize it. Across 4 studies, the authors found that participants scoring in the bottom quartile on tests of humor, grammar, and logic grossly overestimated their test performance and ability. Although their test scores put them in the 12th percentile, they estimated themselves to be in the 62nd. Several analyses linked this miscalibration to deficits in metacognitive skill, or the capacity to distinguish accuracy from error. Paradoxically, improving the skills of participants, and thus increasing their metacognitive competence, helped them recognize the limitations of their abilities.