THIS just might be the most useful book any new postgraduate contemplating
research could ever buy. Research Methods is packed with practical,
common-sense advice about everything from sources of funding and statistical
methods, to writing up your thesis.

Much of this is at the level of the blindingly obvious—with hindsight,
that is. In practice, it is likely to be a lifeline for someone struggling to
get to grips with a new way of life.

Other advice shows sophisticated insight into such things as sampling
methods. And some consists of brutal examples of how not to do research. I
particularly enjoyed Tony Greenfield’s own contribution to methodology,
reporting a true story about what happened when he was asked what could be made
of a medical study in which there were only 32 subjects in a survey, instead of
the recommended 150, and only 7 of them turned out to meet the criteria of the
study. The brutal truth in this case is that the study was worthless.

The point hammered home throughout the book is the need for honesty in
research— including honesty with yourself—and avoiding wishful
thinking.

HAVING recently returned to research (only part time, but still) after a long
spell as a journalistic poacher of research done by others, I fell upon Howard
Becker’s book with cries of delight. Subtitled How to think about your
research while you’re doing it, it looked like just the thing I needed to
get up to speed.

I was sadly mistaken. The research Becker is concerned with is social science
(which I thought, and still tend to think after reading this book, is something
of an oxymoron). The tricks of the trade come, we are told, from “a variety of
fields such as art history, anthropology, sociology, literature and philosophy”.
Yes indeed, and no doubt the anthropologists are embarrassed by the company they
are keeping here.

The fact that the book turned out to be of absolutely no use to me at all in
my contribution to a new determination of the age of the Universe does not,
though, mean that it is not sorely needed by the audience to which it is
addressed.

But this is not a cause for rejoicing in the ranks. In one telling example of
the standards of “scientific” investigation used by the social scientists,
Becker describes a project in which the “scientists” analysed the social status
of the participants in weddings announced in The New York Times
each June over several years. One of their portentous conclusions was
that no Jewish weddings were announced in their sample. They hadn’t realised
that June always falls within one of two Jewish holy periods when religious Jews
do not get married. Such pathetic inability to check the facts would have been
disgraceful during my time as a journalist, let alone as a scientist—yet
Becker finds it necessary to hold the example up as a warning to social
scientists in the 1990s.

It’s a great shame that Richard Feynman has not lived to see this book, and
to offer his comments on it—were he to have been bothered. One of the
anecdotes Feynman liked to tell involved a colleague of his who was running maze
experiments with rats, to test a hypothesis about what made the rats turn left
or right at a T-junction.

Feynman, out of the kindness of his heart, helped out with the statistics.
The research didn’t show what the experimenter was looking for but after the
experiment had been carried out, the colleague returned in great excitement,
having discovered a run in which the rat had alternated its choice at every
junction of the maze—first left, then right, then left, and so on. Surely
the odds against this must be huge?

Patiently, Feynman tried to explain that the way to test the hypothesis that
rats do (or don’t) alternate left and right turns is to formulate the hypothesis
first, then carry out the experiment(s). You can’t choose one odd result from a
pile of data and claim it has particular significance. But his colleague, we are
told, never did get the point.

The tricks that Becker describes are on the same level. They are things that
real scientists learn in the cradle, and use by instinct when doing the
research. Judging by this book, any physicist threatened by cuts in funding
ought to consider a career in the social sciences, where it ought to be possible
to solve the problems the social scientists are worked up about in a trice. Or,
just maybe, to show that they are talking a lot of hot air. Either way, that
would release funding which could then be used to counter the cuts in
physics.

Both sets of calculations were based on Isaac Newton’s theory of gravity. We
now know that the small difference between the predicted and actual orbit of
Mercury shows the beginning of the breakdown of Newtonian theory in strong
gravitational fields. Albert Einstein’s general theory of relativity
spectacularly provided a theory of gravity that goes beyond Newtonian theory,
precisely predicting the observed orbit of Mercury. Little point, you might
think, in writing about the search for Vulcan.

Richard Baum and William Sheehan just about prove that assumption wrong.
In Search of Planet Vulcan (Plenum, $28.95, ISBN 0 306 45567 6)
begins with Greek speculations about the nature of the planets, zips through the
story of Newton’s contribution without adding anything to the traditional
version, and runs through the discoveries of the outer planets.

The meat of the book is the story of Le Verrier—not just his ideas
about Vulcan, but the discovery of Neptune (with credit given to the parallel
work of John Couch Adams). We are almost halfway through before we get to the
story of Vulcan, which is enlivened by discussion of the claims that the
hypothetical planet had been observed, a triumph of expectation over fact which
is a salutary lesson to all scientists.

The story is timely as astronomers spot planets orbiting other stars, using
the techniques that Le Verrier used to probe the then unseen parts of the Solar
System.

Clarity and a refusal to be fooled by seductively offered, but insecurely
founded, ideas are the hallmarks of Trefil’s work. Doubtless this is a tribute
to his background in physics. He provides one of the most thorough and
persuasive critiques of the often extravagant claims about the ability of other
species, notably chimpanzees, to use language that I have seen. He also provides
a clear summary of how evolution works. And he makes the point that the
possibility that human intelligence is a unique attribute does not make us any
more special than the fact that the octopus, for example, is unique. Being
unique does not place us outside evolution, or outside the scope of scientific
investigation. This is said with more force than seems necessary, but perhaps he
is responding to the claims of creationists in the US.

At the heart of Are We Unique is Trefil’s explanation of why the
human brain is not the same as an electronic computer, and his persuasive
argument that our kind of intelligence and consciousness has to do with the
physical complexity of our brains. The key point here is that there may well be
a threshold of complexity above which consciousness kicks in, so that even if
the brain (specifically the cerebral cortex) of another species has, say,
one-hundredth of the complexity of a human brain, that does not mean that an
individual member of that species has 1 per cent of our consciousness.
Consciousness is arguably an all-or-nothing phenomenon that switches on at some
level of complexity.

It is this kind of discussion that leads Trefil to the conclusion that our
form of intelligence is not the same as machine intelligence. I would agree with
this up to a point. But Trefil fails to address the fact that if improvements in
computing continue at the present rate for a few more decades, it will be
possible to build a machine that simulates the way a human brain works, right
down to the neurons, with all the complexity of a human brain. It might be
expensive and wasteful of computer power, like the simulator I can run on my
PowerMac that makes it pretend it is a ZX Spectrum. I see no reason, however,
why such a virtual brain should not achieve exactly the same kind of
consciousness as the human brain, even if you will never manage to do this
simply by adding more RAM to your home PC.

But in all probability I would never have thought of this if Trefil had not
led me by the hand through this version of what he understands by consciousness,
and the major role of complexity. It is a sign of just how good this book is
that about the only nit I can find to pick is that, although eclectic enough to
include a discussion of Isaac Asimov’s three laws of robotics, it consistently
misspells the name of his robot detective, R. Daneel Olivaw. And if that is the
worst mistake in the book, you can imagine how good the rest of it is.

Rabi Mohapatra and his colleagues at the University of Maryland in College
Park followed up an idea that dates back to the 1980s. This was that every
elementary particle has a partner in a mirror world, an idea that would help
tidy up theories of how the four forces of nature acted as one just after the
Universe was born.

The two components, each made up of its own set of fundamental
particles—such as the electron and the three normal types of
neutrino—would be mixed up. The force of gravity could easily clump all
the mirror matter into mirror stars and mirror planets. But that would be the
only force of our world that they would feel. Mirror matter would not feel the
electromagnetic force, so mirror stars would be invisible.

Mohapatra realised that if one of the three neutrinos that would exist in the
mirror world was much heavier than its everyday counterpart, this could explain
why the Sun’s nuclear reactions produce only a third of the neutrinos that
theory predicts.

Physicists have suggested that the reason for this is that the type of
neutrino expected from nuclear reactions is turning into the two other normal
types—which are as yet undetectable—en route to the Earth. This
happens through a so-called oscillation—two or three neutrinos are
effectively “mixed up” and flip from one type to another over a regular
distance. Some inconclusive evidence for this comes from detectors at Los Alamos
National Laboratory in New Mexico (see “Cosmic changelings”, New
żěè¶ĚĘÓƵ, 16 March 1996, p 28).

The oscillation distance of the neutrino mixture depends on the masses of the
neutrinos involved. The masses of everyday neutrinos seem too small to allow
oscillations over the distances that would explain the solar neutrino problem.
But a fourth, heavier neutrino from the mirror universe could interact with the
normal neutrinos through gravity and make the oscillation distance just right,
says Mohapatra.

Now, working with his colleague Vigdor Teplitz of the University of Maryland,
Mohapatra has looked in detail at the implications of this theory for the mirror
universe (Astrophysical Journal, vol 478, p 29). If the mirror world
contains a heavy neutrino, it turns out that beta-decay—which involves
neutrinos—is much faster there. In that case, mirror neutrons would decay
so quickly that the only element that could exist is mirror hydrogen.

Without neutrons, nuclear fusion would be impossible, so there would be no
stars, says Mohapatra. As the Universe expands, the mirror hydrogen would clump
into “globs” the size of dense star clusters. The globs would form
part—perhaps all—of the invisible “dark matter” that exerts a pull
on stars and galaxies. Mohapatra adds that the idea can be tested, because the
globs would bend the light from background galaxies and focus it just like a
lens.

Rocky Kolb of Fermilab, near Chicago, calls Mohapatra’s work “a really nice
study”, and adds that it could help to extend the current Standard Model of
particle physics. “There seem to be several loose threads fraying at the edges
of the Standard Model, and nobody is sure which one to pull to make the model
unravel,” he says. “Perhaps this is it.”

In 1995, Stephen Belcher of the University of Reading and his students
calculated how much water falls on top of your head and how much you sweep up on
your front as you move forward (Weather, vol 50, p 367). Obviously, you
would get wettest standing still, and less wet the faster you moved. But the
Reading team found that the benefits of running faster than about 3 metres per
second—which they described as a walking pace—were tiny.

Thomas Peterson and Trevor Wallis, meteorologists at the National Climatic
Data Center in Asheville, North Carolina, had a hunch that this was wrong. They
realised that the Reading team had overestimated the average walking pace, so
they reworked the calculations for a walking pace of 1.5 metres per second and a
running speed of 4 metres per second.

Peterson and Wallis conclude in the latest issue of Weather (vol 52,
p 93) that a walker would get 16 per cent wetter than a runner over a distance
of 100 metres in drizzle. In heavy rain, this would rise to 23 per cent. When
the researchers allowed for the way that runners tend to lean forward,
sheltering the front of their bodies but increasing the rainfall on their backs,
they found that a walker would get 36 per cent wetter than a runner in heavy
rain.

Not content with theory alone, Peterson and Wallis decided to test their
ideas. “If verification requires an $80 million satellite, one may have
to forgo verification,” says Peterson. “But if it involves a simple experiment,
that’s another matter.” Peterson and Wallis are roughly the same size. Wearing
identical clothing, one ran 100 metres in heavy rain and the other walked.

They weighed their clothes before and after the experiment. This showed that
the walker had absorbed 0.22 kilograms of water, while the runner had soaked up
only 0.13 kilograms. This is about 40 per cent less, in line with the model’s
predictions.

Belcher says that his team’s work was a bit of fun, and that apart from the
confusion over what a typical walking speed is, their results were similar to
those of Peterson and Wallis. “I’m delighted to see that their experiments gave
results in qualitative agreement with the model,” says Belcher.

But why not just take an umbrella? For anyone thinking of taking the easy way
out, Wallis has a warning: “Running with an umbrella has a negative impact on
your aerodynamics.”

Now a detailed study of waves on the star’s surface has revealed almost
perfect agreement with models of the Sun’s structure. The standard physical
description of the Sun seems to be safe. It is the particle physicists and their
ideas about neutrinos who need to revisit the drawing board.

The temperature at the Sun’s core is an important piece of the solar neutrino
puzzle. The core would generate electron neutrinos at the detected
rate—about one-third of that predicted by models—if it was even 10
per cent cooler than physicists believe. But physicists cannot directly measure
core temperature to that accuracy.

Instead, John Bahcall of the Institute for Advanced Study in Princeton, New
Jersey, and colleagues elsewhere in the US and in Denmark used new measurements
of the waves that appear on the Sun’s surface. The team determined, based on the
waves’ trajectories, which of these would pass through the core of the Sun.
These revealed its temperature and structure rather as details of the interior
of Earth can be gleaned by studying how earthquake waves pass through it. Their
observations match the standard solar model to an accuracy better than 0.2 per
cent (Physical Review Letters, vol 78, p 171). “I find the precise
agreement just stunning,” says Bahcall.

This dramatic achievement shows that electron neutrinos must be disappearing
before they hit Earth. Some physicists have already speculated that they might
convert to other flavours of neutrinos—the muon and tau (“Cosmic
changelings”, żěè¶ĚĘÓƵ, 16 March 1996, p 28).

John Ellis, a particle physicist with CERN, accepts that the next step is up
to those in his discipline. “The solar neutrino problem cannot be solved by
tinkering with the standard model of the Sun.”