Uncommon Descent Serving The Intelligent Design Community

Does an Element of Subjective Judgment Exclude a Research Program from the Realm of “Science”?

Share
Facebook
Twitter
LinkedIn
Flipboard
Print
Email

Congratulations to all of those Darwinists who seek to exclude ID from science whenever the CSI in a structure or DNA sequence is difficult to quantify exactly.  You’ve just excluded a highly influential form of evolutionary analysis (cladistics) from science as well.  The following lengthy quote is from Adrain, Jonathan M.; Edgecombe, Gregory D. & Lieberman, Bruce S., Fossils, Phylogeny, and Form: An Analytical Approach, New York: Kluwer Academic (2002), pp 56-57:

Phylogenetic inference is pivotal to an understanding of the systematics of any group.  Cladistics offers an objective framework for the analysis of data that inevitably incorporates elements of subjectivity (Hennig 1966, Swofford 1993).  A cladogram is a hypothesis of relationships derived from a set of putatively homologous morphological and/or molecular characters (Forey 1992), to which is added information on character polarity or the nature of an outgroup.  If homologous organs or characters are defined as those jointly inherited from a common ancestor (Simpson 1961, Hennig 1966), it becomes impossible to identify homologies without access to the true phylogeny (a problem of circularity:  Jardine and Sibson 1971).  Hence, criteria of compositional and structural similarity are used in practice.  Compositional similarity refers to resemblance in terms of biological or chemical constituents (the composition of the organs).  Structural correspondence refers to the spatial or temporal arrangement of parts, structure of biochemical pathways, or the sequential arrangement of organized structures (Sneath and Sokal 1973).  The number of potential characters is limited only by our ability to recognize putative homologies at increasingly fine scales.

Inevitably, even the most rigorous tests of homology can fail to identify character states that are similar because of convergence or ‘reversal’ (‘homoplasy’, rather than direct, common descent).  Most real data sets therefore contain character conflict (Strauch 1984, Deleporte 1993).  This is usually resolved using some optimality criterion (e.g., parsimony) to derive one or more cladistics hypotheses (which will reject some fraction of the supposed homologies).

Various types of data and analytical techniques are employed in cladistics, sometimes yielding widely differing results (Wiley 1981).  Nonetheless, there is consensus on the nature of the pattern being sought, and the objective reality of the process that produced it (cladogenesis).  There is only one true evolutionary tree, and the diversity of approaches therefore all have the same ultimate goal (Wilkinson 1992).  Inevitably, the process of selecting characters for analysis is subjective, and amounts to a radical form of character weighting (Meacham 1994, Wilkinson 1994a).  The sample is also likely to be biased towards more obvious features, and frequently towards those with some form of historical precedence (Pearson 1999).

The absence of complete objectivity at all stages in a cladistics analysis in no sense detracts from its value in producing hypotheses of relationships.  Other (non-cladistic) approaches in systematics also operate on finite data sets and incorporate similar assumptions.  Often, these do not produce hypotheses directly, but serve to describe aspects of the data, frequently offering additional insights into evolutionary processes (Foote 1996).  All results (phylogenetic and otherwise) should therefore be presented along with the original data and sufficient information to all the analysis to be repeated.

(emphasis mine)

Let’s count up the subjective judgments that go into constructing a cladogram.  I see at least the following (there are almost certainly more).

  1. Which characters am I going to select for analysis?
  2. Are these structures homologous?
  3. Is there “resemblance” of biological or chemical constituents?
  4. Are the spatial and temporal arrangement of parts similar?
  5. Are the character differences upstream or downstream?
  6. Homology or Homoplasy?
  7. Is there “structural correspondence”?

No wonder different scientists’ analyses yield “wildly differing results.”

Consider the following sentence extracted from the quotation above:

The absence of complete objectivity at all stages in a cladistics analysis in no sense detracts from its value in producing hypotheses of relationships.

Is the following a fair extrapolation of the authors’ logic:

The absence of complete objectivity at all stages in a ID analysis in no sense detracts from its value in producing hypotheses of design.

Comments
BTW, please do not misunderstand the OP. I am not saying that CSI is never objectively measurable. It is. See Joe's comment at 7.Barry Arrington
November 4, 2014
November
11
Nov
4
04
2014
08:00 AM
8
08
00
AM
PDT
Robb:
I’d be happy with any kind of quantitative result, be it a rough estimate, a range, a lower limit (e.g. “at least 500 bits”), or whatever.
Have you asked a cladistics scientist to give you any kind of quantitative result, be it a rough estimate, a range, a lower limit, or whatever” of the degree of “similarity” of two putatively homologous character traits?Barry Arrington
November 4, 2014
November
11
Nov
4
04
2014
07:57 AM
7
07
57
AM
PDT
Hi R0bb, read the Durston, et al., paper: Information means here the precise determination of sequence, either of bases in the nucleic acid or on amino acid residues in the protein. Each protein consists of a specific sequence of amino acid residues which is encoded by a specific sequence of processed mRNA. Each mRNA is encoded by a specific sequence of DNA.  The point being is biological information refers to the macromolecules that are involved in some process, be that transcription, editing, splicing, translation and functioning proteins. No one measures the biological information in a random sequence of DNA nor any DNA sequence not directly observed in some process. The best one can do with any given random DNA sequence is figure out its information carrying capacity. You couldn't tell if it was biological information without a reference library. And Leslie Orgel first talked about specified complexity wrt biology:
In brief, living organisms are distinguished by their specified complexity. Crystals are usually taken as the prototypes of simple well-specified structures, because they consist of a very large number of identical molecules packed together in a uniform way. Lumps of granite or random mixtures of polymers are examples of structures that are complex but not specified. The crystals fail to qualify as living because they lack complexity; the mixtures of polymers fail to qualify because they lack specificity.
As far as I can tell IDists use the terms in the same way. Dembski and  Meyer make it clear that it is sequence specificity that is central to their claims. That is the whole point- if sequence specificity matters the tighter the specification the less likely blind physical processes could find it. Yup those dreaded probabilities again, but seeing yours doesn't come with a testable model it's all we have. See Is Intelligent Design Required for Life? With that said, to measure biological information, ie biological specification, all you have to do is count the coding nucleotides of the genes involved for that functioning system, then multiply by 2 (four possible nucleotides = 2^2) and then factor in the variation tolerance: from Kirk K. Durston, David K. Y. Chiu, David L. Abel, Jack T. Trevors, Measuring the functional sequence complexity of proteins, Theoretical Biology and Medical Modelling, Vol. 4:47 (2007):
[N]either RSC [Random Sequence Complexity] nor OSC [Ordered Sequence Complexity], or any combination of the two, is sufficient to describe the functional complexity observed in living organisms, for neither includes the additional dimension of functionality, which is essential for life. FSC [Functional Sequence Complexity] includes the dimension of functionality. Szostak argued that neither Shannon’s original measure of uncertainty nor the measure of algorithmic complexity are sufficient. Shannon's classical information theory does not consider the meaning, or function, of a message. Algorithmic complexity fails to account for the observation that “different molecular structures may be functionally equivalent.” For this reason, Szostak suggested that a new measure of information—functional information—is required.
The following is interesting:
 First, as observed in Table ?Table1,1, although we might expect larger proteins to have a higher FSC, that is not always the case. For example, 342-residue SecY has a FSC of 688 Fits, but the smaller 240-residue RecA actually has a larger FSC of 832 Fits. The Fit density (Fits/amino acid) is, therefore, lower in SecY than in RecA. This indicates that RecA is likely more functionally complex than SecY.  (results and discussion section) 
Joe
November 4, 2014
November
11
Nov
4
04
2014
07:52 AM
7
07
52
AM
PDT
I don't even need a positive number. All I need is someone to show me how Glacier or Sand (as asserted by a ID proponent)has zero or null dFSCI or any variant thereof.Me_Think
November 4, 2014
November
11
Nov
4
04
2014
07:42 AM
7
07
42
AM
PDT
Barry:
Congratulations to all of those Darwinists who seek to exclude ID from science whenever the CSI in a structure or DNA sequence is difficult to quantify exactly.
Which Darwinists have required exact quantities? Can you name names? I'd be happy with any kind of quantitative result, be it a rough estimate, a range, a lower limit (e.g. "at least 500 bits"), or whatever. If that's asking too much, then maybe you can show us an example of someone obtaining a binary result of "it has CSI" or "it doesn't have CSI". The important thing is the method employed to obtain the result, i.e. the operational definition. Without this, the challenge to find a natural process that produces CSI is literally meaningless. I'll give up on trying to get a quantitative operational definition (despite Dembski's plea to "Do the calculation. Take the numbers seriously."), and I'll try to get a binary operational definition. Can you fill in the following blanks: According to ______________'s assessment, _____________ has CSI [or doesn't have CSI]. Here is a reference that shows how they obtained this result: _________________________. Once we know the method, we can apply it to other things in an attempt to meet your challenge. Thanks in advance, Barry.R0bb
November 4, 2014
November
11
Nov
4
04
2014
07:38 AM
7
07
38
AM
PDT
I haven’t seen that argument made,
It is made incessantly on these pages. It is implicit every time an anti-ID poster insists that ID has nothing to say unless the CSI in a particular feature can be calculated with precision.Barry Arrington
November 4, 2014
November
11
Nov
4
04
2014
07:30 AM
7
07
30
AM
PDT
Are you therefore prepared to reject the assertions of all of those who call for ID to be cashiered from the ranks of science if there is even a whiff of subjectivity?
Of course. I haven't seen that argument made, but bad arguments are bad arguments. . . . and yes cladistics based on morphology is particularly prone to this.
Understatement. “These two characters look kinda alike” is about as subjective as it gets.
It is more than that. The people who make these comparisons have spent a lot of time looking at specimens, and the traits they chose have to be identifiable by other people (after all, part of the reason to do this is so that other people can identify species, for example with a key). The characters thus do have to be spelled out. So whilst there will be some subjectivity, taxonomists try to reduce it, because the alternative is to make their work useless for other people.
Sure, that is also one of the points of the quoted text. If all of the subjective decisions are laid bare, the resulting cladogram falls out no matter who is constructing it. That does not make the decisions that result in the cladogram in the first place less subjective.
Indeed, which is why it would be interesting to see such an analysis of a design inference.Bob O'H
November 4, 2014
November
11
Nov
4
04
2014
07:24 AM
7
07
24
AM
PDT
Thanks for your comment Bob.
I think most people who study how science is done now accept that there is subjectivity, . . .
Are you therefore prepared to reject the assertions of all of those who call for ID to be cashiered from the ranks of science if there is even a whiff of subjectivity?
. . . and yes cladistics based on morphology is particularly prone to this.
Understatement. “These two characters look kinda alike” is about as subjective as it gets.
One point . . .
Sure, that is also one of the points of the quoted text. If all of the subjective decisions are laid bare, the resulting cladogram falls out no matter who is constructing it. That does not make the decisions that result in the cladogram in the first place less subjective.Barry Arrington
November 4, 2014
November
11
Nov
4
04
2014
07:06 AM
7
07
06
AM
PDT
I think most people who study how science is done now accept that there is subjectivity, and yes cladistics based on morphology is particularly prone to this. One point: the decision about whether similarity of states is homology or homoplasy is not itself subjective, as the point of using methods like parsimony is to estimate one tree, and from that one can identify homoplasies. Of course, there is subjectivity in what traits are chosen, but if these are all laid out clearly then they can be checked, so someone with a different opinion can see if they get a different result. I'd be interested to see a similar analysis of an ID analysis that produced a hypothesis of design. In part because it would help to identify the parts of ID that would need more work. A big advance in systematics was the introduction of parsimony, as it was a tool for making trees that was less subjective than what had gone before. Similarly, using DNA to create phylogenetic trees removed a lot of the problems with choosing and coding states.Bob O'H
November 4, 2014
November
11
Nov
4
04
2014
06:44 AM
6
06
44
AM
PDT
1 2 3

Leave a Reply