Some creationists have claimed that historical measurements of the speed of light ("C") show a decrease of half a percent across the last 300 years. By drawing a curve (rather than a straight line) through a graph of measured values, they produce a curve which goes to infinite speed at 6,000 years ago.
The hope is to reconcile a young earth with the size of the universe. Astronomers measure distant galaxies as being billions of light-years away. But, if C used to be a million times higher, then light could travel that far in only thousands of years. (An alternative Young Earth argument is that the light was created in transit.)
Strangely, the curve says that the speed of light stopped changing somewhere around 1960, and is now constant. I do not know of a Creationist explanation of why this happened at the exact moment when scientists built extremely accurate measurement equipment.
Barry Setterfield suggested that the speed of light was decreasing. He has written about this a number of times, for example
The Velocity of Light and the Age of the Universe, Part 1, Ex Nihilo vol. 4, no. 1, 1981The idea that light had a finite speed was suggested by Romer (or Roemer) in 1675. He used an astronomical technique to measure its speed, and his measurement was very inaccurate by modern standards. Setterfield has catalogued over 160 subsequent measurements, which got more and more precise with time. Setterfield claims that older measurements gave too-high values of C suspiciously often, and that the measurement errors do not completely account for that.
Setterfield argued that some of the measurements should be included in his graph, and some rejected. In some cases, he used adjusted values, rather than the values published by the original authors. For his curve, he used the highly unusual equation
log C = A + B log sin (90/6000 t )where A is about 5.47, B is about -1.94, and t = years since Creation. He assumed that Creation was 6000 years ago. The 90/6000 is there so that today, at t=6000, we are taking the sine of 90 degrees. In later papers, the equation was changed to use a cosec2.
The Institute for Creation Research (ICR) addressed the issue in their monthly monograph:
Has the Speed of Light Decayed?, Gerald Aardsma, Impact May 1988Aardsma, an ICR member, wished to "caution creationists against a wholesale, uncritical acceptance of the Norman and Setterfield hypothesis." He noted, for instance, that Setterfield had completely ignored the error bars and treated every data point as having equal significance. He noted that a more standard treatment of the same data indicated no measurable change at all in the speed of light.
Aardsma also had some issues with the data adjustments. Setterfield used an adjusted figure for Romer's measurement in 1675, which brought it up from being 24% too low to being 1.5% high. But he used an unadjusted figure for Cassini's much higher 1693 measurement.
Hugh Ross, an astronomer and Old-Earth Creationist, said in Facts and Faith (Summer 1989) that the whole issue was an abuse of Christian apologetics. He pointed out that if C was as little as 3% higher, it would wipe out all life on earth, by making the Sun hotter.
The idea has not been abandoned. For example, in Creation Ex Nihilo vol. 12, no. 1, 1998, Creationist Malcolm Bowden argues for the idea. However, his article admits "There are more creationists who have written against [speed of light decay] than for it."
Bowden does make the misstatement that the idea was "promoted by the Stanford Research Institute". Although a report does exist that is on SRI stationery, it was never sanctioned by SRI.
The talk.origins archive agrees with Aardsma and Ross that Setterfield mishandled data. In particular, Setterfield took the adjusted Romer value as 0.5% higher than today's C value. But the article he took the adjustment from, says that the discrepancy is less than 0.5%.
Creationist Walter Brown argues that if C has decreased by a millionfold, then faraway objects are being seen in very, very slow motion. His argument seems reasonable. However, there are astronomical objects which pulsate at more or less known frequencies - specifically, pulsars and Cepheid variable stars. The frequencies of these objects are known to be independent of distance. Cepheids near us in the Milky Way run at the same speed as Cepheids in other galaxies. Therefore, by Brown's argument, C is constant.
The scientific community in general has ignored Setterfield. Adjustments to Romer's value are in the scientific literature precisely because the constancy of physical law is much studied. The historical measurements of C are not the only kind of evidence, nor are they the most "powerful" kind.