Tom - Mars Reconnaissance Orbiter question

[tvf]
"This prediction process to test the significance of statistical anomalies is standard science, and is known as "the a priori principle". Here's another example of its application. In 2001, JP Levasseur found <<what looks like an image of part of an animal>> in an MGS photo, and wondered if it was natural or perhaps another example of flat art ...

So he set up an a priori test:
1) If natural, then when future images reveal what's to the right of the strip, it will be plain sand, craters, hills, mesas, irregular lava flows, or countless other natural landforms.
2) But if artificial, the landform will be very specific -- the completion of the animals hind quarters, missing leg, and a tail of the right proportions, uniform shape, orientation, and location with respect to the animal's body."

<Tom's last question>
"Do you agree that, when the new data is seen, this will be a good test?"


[jrich]
"I think you can guess what I think of the *flat art* phenomena."

===

Although I share your skepticism on the flat art thing, I would like to hear your answer to Tom's last question. (Note that he asks if it would be a good test, not if it would be a perfect test.)

LB

<blockquote id="quote"><font size="2" face="Verdana, Arial, Helvetica" id="quote">quote:<hr height="1" noshade id="quote"><i>Originally posted by neilderosa</i>
<br />I get the distinct impression that jrich is not really arguing science, but epistemology. How do we know what we know?

Neil

<hr height="1" noshade id="quote"></blockquote id="quote"></font id="quote">

Has jrich ever stated what would convince him? That's really the issue, isn't it? I, for one, would be interested in hearing it, since he appears to be saying that there is NO objective test that could be designed. How about some subjective threshold that we might cross. Anything?

rd

I think the test is a good one, of course, but more important is what was the result? This is where things seem hazy to me. I found the website I remember seeing over a year ago, with the original "puma" image and the prediction for a tail. It also includes supposedly subsequent photos that the authors claim does show a tail, and thus adds weight to the artificiality hypothesis because this is what was predicted. HOWEVER, I think the results are ambiguous, if not outright negative. First, here's the link to the website:

marsartifacts.tripod.com/lioness.html

The authors claim that the original "puma" image was published by JPL prior to *two* subsequent strips appearing to show a tail. However, the second, higher reolution image to me doesn't show a tail anywhere. This is suspicious to me. They claim there is a tail in both the new images and there simply isn't one at all in the much-clearer, higher resolution image. HEre they are:



Now in the lower-resolution image onthe left (THEMIS I02417005) there is what looks like a tail, but in the higher resultion image on the right (MOC S050025) which is the very adjascant strip, there is no such thing. Yet the authors don't acknowledge this. A case of trying to prove their bias?

Emanuel

<blockquote id="quote"><font size="2" face="Verdana, Arial, Helvetica" id="quote">quote:<hr height="1" noshade id="quote"><i>Originally posted by emanuel</i>
<br />I think the test is a good one, of course, but more important is what was the result?<hr height="1" noshade id="quote"></blockquote id="quote"></font id="quote">In part, this was a test of whether JR (and other readers of like mind) understood and accepted the methodology as valid. That was more important here than the test results. I do not consider this case as strong evidence of artificiality. To me, the result is more evidence of functionality, given proved artificiality elsewhere. But I find it hard to understand why anyone would not see this as legitimate methodology for testing artificiality, at least in principle.

BTW, your second link did not work properly with this software. But it can be found near the page end at the first link.

<blockquote id="quote"><font size="2" face="Verdana, Arial, Helvetica" id="quote">quote:<hr height="1" noshade id="quote">Now in the lower-resolution image onthe left (THEMIS I02417005) there is what looks like a tail, but in the higher resultion image on the right (MOC S050025) which is the very adjascant strip, there is no such thing.<hr height="1" noshade id="quote"></blockquote id="quote"></font id="quote">It's there, and pretty well-defined. But you may need to adjust the contrast to see it on your monitor. It was also invisible on my monitor until I put it into image software and adjusted the contrast, which made it obvious. -|Tom|-

<blockquote id="quote"><font size="2" face="Verdana, Arial, Helvetica" id="quote">quote:<hr height="1" noshade id="quote"><i>Originally posted by jrich</i>
<br />The tests rely on human beings finding certain patterns in images of poor quality relative to what is really necessary for the task.<hr height="1" noshade id="quote"></blockquote id="quote"></font id="quote">?? The image quality problems have long since been corrected.

<blockquote id="quote"><font size="2" face="Verdana, Arial, Helvetica" id="quote">quote:<hr height="1" noshade id="quote">Worse still, those implementing the tests and searching for more detailed features, like eyebrows and nostrels, are the same people who are seeing gross features in the Face images - the eye sockets, nose, mouth - that many people, like myself, simply don't see with the degree of certainty that they do.<hr height="1" noshade id="quote"></blockquote id="quote"></font id="quote">It was very hard to make any sense of the 1998 Face image until we had the spacecraft and Sun data and were able to relate it, feature-by-feature, to the Viking images that preceded it. The DeRosas used a good analogy above. With too much magnification, you can't tell what part of the Mona Lisa you are looking at.

Your statements here, especially this last one just quoted, left me wondering how you could possible be saying that. Then I realized you must never have looked at, or are unable to download, animation #1 at
metaresearch.org/media%20and%20links/animations/animations.asp
Am I correct about that? To me, the animation answers all questions about any suggestion of ambiguity in the MGS raw image.

To me, it also renders all your other protests moot. Have a look if you can, and tell me what ambiguity remains after seeing that. -|Tom|-

<blockquote id="quote"><font size="2" face="Verdana, Arial, Helvetica" id="quote">quote:<hr height="1" noshade id="quote"><i>Originally posted by Larry Burford</i>
<br />[tvf]
"This prediction process to test the significance of statistical anomalies is standard science, and is known as "the a priori principle". Here's another example of its application. In 2001, JP Levasseur found &lt;&lt;what looks like an image of part of an animal&gt;&gt; in an MGS photo, and wondered if it was natural or perhaps another example of flat art ...

So he set up an a priori test:
1) If natural, then when future images reveal what's to the right of the strip, it will be plain sand, craters, hills, mesas, irregular lava flows, or countless other natural landforms.
2) But if artificial, the landform will be very specific -- the completion of the animals hind quarters, missing leg, and a tail of the right proportions, uniform shape, orientation, and location with respect to the animal's body."

&lt;Tom's last question&gt;
"Do you agree that, when the new data is seen, this will be a good test?"


[jrich]
"I think you can guess what I think of the *flat art* phenomena."

===

Although I share your skepticism on the flat art thing, I would like to hear your answer to Tom's last question. (Note that he asks if it would be a good test, not if it would be a perfect test.)

LB
<hr height="1" noshade id="quote"></blockquote id="quote"></font id="quote">
I looked at Mr. Levasseur's website. Apparently there are newer images that show the missing part of the "animal". Mr. Levasseur claims to see a "tail", but I didn't see anything even with the aid of his outlines. I did see some other things though that were missing from the original image. The new image shows the context of the "animal" in relation to the surrounding landscape. It appears that the "animal" is near the edge of a wadi or perhaps even within it (the image doesn't show enough to be sure). So rather than expecting to find "plain sand, craters, hills, mesas, irregular lava flows, or countless other natural landforms" in the missing part of the image if it were natural, one would actually have expected to find features consistent with a dry riverbed, the same sort of features that make up the visible part of the "animal". I think this illustrates the real problem with any of these natural vs. artificial tests: the features can satisfy tests for both artificiality and naturalness at the same time.

Tom is used to analyzing astronomical images, but this type of image analysis is not like that which is done in astronomy. He and others are not analysing images and claiming to see real faces and animals, but artistic representations of them. Not real objects, but <b>abstract</b> images of real objects, and apparently poor ones at that. As I said in a previous post, I liken this to a Rorschach test with images of Mars as the inkblots. Everyone sees many things in the inkblots of a Rorschach test, but there is no artistic intent in the blots of ink. They are all "natural".

JR