A hundred percent conclusive science

Estimation and certainty in Maisie's galaxy


Keith S. Taber


An image from the James Webb Space Telescope
(released images are available at https://webbtelescope.org/resource-gallery/images)

NASA's James Webb Space Telescope is now operational, and offering new images of 'deep space'. This has led to claims of finding images of objects from further away in the Universe, and so from further back in time, than previously seen. This should support a lot of new scientific work and will surely lead to some very interesting findings. Indeed, it seems to have had an almost immediate impact.

Maisie's galaxy

One of these new images is of an object known as:

CEERSJ141946.35-525632.8

or less officially (but more memorably) as

Masie's galaxy.

A red smudge on one of the new images has been provisionally identified as evidence of a galaxy as it was less than 300 000 000 years after the conjectured 'big bang' event understood as the origin of the universe. The galaxy is so far away that its light has taken since then to reach us.

Three hundred million years seems a very long time in everyday terms, but it a small fraction of the current age of the universe, believed to be around fourteen billion years. 1

300 000 000 years

≪ 14 000 000 000 years

The age estimate is based on the colour of the object, reflecting its 'redshift':

"Scientists with the CEERS Collaboration have identified an object–dubbed Maisie's galaxy in honor of project head Steven Finkelstein's daughter–that may be one of the earliest galaxies ever observed. If its estimated redshift of 14 is confirmed with future observations, that would mean we're seeing it as it was just 290 million years after the Big Bang."

University of Texas at Austin, UT News, August 04, 2022

(CEERS is the Cosmic Evolution Early Release Science Survey.)

This finding is important in understanding the evolution of the universe – for example, observing the earliest galaxies puts a limit on how long the universe existed before star formation started. (Although the episode was called 'The first galaxies at the universe's dawn' Masie's galaxy is thought to contain heavier elements that were produced in even earlier stars.)

Uncertainty in science (and certainty in reporting science)

So, the claim is provisional. It is an estimate awaiting confirmation.

Strictly, science is concerned with provisional knowledge claims. This is not simply because scientists can make mistakes. All measurements are subject to limits in precision (measurement 'errors'). More fundamentally, all measurements depend on a theory of the instrumentation used, and theoretical knowledge is always open to being revisited on the basis of new ways of understanding.

We may not expect the theory behind the metre rule to change any time soon (although even there, our understanding shifted somewhat with Einstein's theories) but many scientific observations depend on highly complex apparatus, both for data collection and analysis. Despite this, science is often represented in the media, both by commentators and sometimes scientists themselves, as if it produced absolute certainty.

Read about science in public discourse and the media

Read about scientific certainty in the media

A rough estimate?

In the case of Maisie's galaxy, the theoretical apparatus seems to be somewhat more sophisticated than the analytical method used to provisionally age the object. This was explained by Associate Professor Steve Finkelstein when he was interviewed on the BBC's Science in Action episode 'The first galaxies at the universe's dawn'.


Masie's galaxy – it's quite red.
The first galaxies at the universe's dawn. An episode of 'Science in Action'

Professor Finkelstein explained:

"We can look deep into out past by taking these deep images, and we can find the sort of faintest red smudges and that tells us that they are extremely far away, and from exactly how red they are we can estimate that distance."

Associate Professor Steve Finkelstein

So, the figure of 290 000 000 years after the big bang is an estimate. Fair enough, but what 'caught my ear', so to speak, was the contrast between the acknowledged uncertainty of the current estimate, and the claimed possibility of moving from this to absolute knowledge,

"If this distance we have measured for Masie's galaxy, of a red shift of 14, holds true, and I can't stress enough that we need spectroscopic confirmation to precisely measure that distance. [*] Where you take a telescope, could be James Webb, could be a different telescope, you observe it [the galaxy] and you split the light into its component colours, and you can actually precisely measure – measure the red shift, measure the distance – a hundred percent conclusively."

Associate Professor Steve Finke
[* To my ear, there might well be an edit at this point – the quote is based on what was broadcast which might omit or re-sequence parts of the interview.]

Spectroscopic analysis allows us to compare the pattern of redshifted spectral lines due to the presence of elements absorbing or emitting radiation, with the position of those lines as they are found without any shift. Each element has its own pattern of lines – providing a metaphorical fingerprint. A redshift (or blueshift) moves these lines to different parts of the spectrum, but does not change their collective profile as all the lines are moved to the same extent.


Spectral lines can be used like fingerprints to identify substances.
(Image by No-longer-here from Pixabay)

Some of these lines are fine, allowing precise measurements of wavenumber/frequency, and there are enough of them to be able to make very confident assignments of the 'fingerprints', and use this to estimate the shift. We might extend our analogy to a fingerprint on a rubber balloon which had been stretched since a fingerprint was made. In absolute terms, the print would no longer (or 'no wider' for that matter) fit the finger that made it, but the distortion is systematic allowing a match to be made – and the degree of stretching to be calculated.


If a pattern is distorted in a systematic way, we may still be able to match it to an undistorted version
(Original images by Clker-Free-Vector-Images (balloon), OpenClipart-Vectors (print) and Alexander (notepad) from Pixabay)

Yet, even though this is a method that is considered well-understood, reliable, and potentially very accurate and precise 2, I am not sure you can "precisely measure, measure the redshift, measure the distance. A hundred percent conclusively". Science, at least as I understand it, always has to maintain some small level of humility.

Scientists may be able to confirm and hone the estimate of 290 000 000 years after the big bang for the age of Maisie's galaxy. Over time, further observations, new measurements, refinement in technique, or even theory, may lead to successive improvements in that age measurement and both greater accuracy and greater precision.2 But any claim of a conclusive measurement to a precision of 100% has a finality that sounds like something other than science to me.


Notes

1 Oddly, most webages I've seen that cite values for the age of the universe do not make it explicit whether these are American (109) or English (1012) billions! It seems to be assumed that, as with sulfur [i.e., sulphur], and perhaps soon aluminum and fosforus, we are all using American conventions.


2 Precision and accuracy are different. Consider an ammeter.


An ammeter (Image by Gerd Altmann from Pixabay)

Due to the method of reading a needle position against a scale there is a limit to precision (perhaps assumed to the nearest calibration, so to ±0.5 calibrations). This measurement error of ±0.5 units is, in effect, a limit in detail or resolution, but not an 'error' in the everyday sense of getting something wrong. However, if the meter had been miscalibrated, or over time has shifted from calibration, so the needle is misaligned (so perhaps the meter reads +0.15 A when it is not connected into a circuit) then that is inaccuracy. There is always some level of imprecision (some limit on how precise we can be), even when we have an accurate reading.


In science, a measurement normally offers a best estimate of a true value, with an error range acknowledging how far the real value might be from that best estimate. See the example below: Measurement B claims the most precision, but is actually inaccurate. Measurement A is the most accurate (but least precise).

If we imagine that a galaxy was being seen as it was

275 000 000 years after the big bang

and three measurements of its age were given as:

A: 280 000 000 ± 30 000 000 years after the big bang

(i.e., 250 000 000 – 310 000 000)

B: 290 000 000 ± 10 000 000 years after the big bang

(i.e., 280 000 000 – 300 000 000)

C: 260 000 000 ± 20 000 000 years after the big bang

(i.e., 240 000 000 – 280 000 000)

then measurement B is more precise (it narrows down the possible range the most) but is inaccurate (as the actual age falls outside the range of this measurement). Of course, unlike in such a hypothetical example, in a real case we would not know the actual age to allow us to decide which of the measurements is more accurate.