The Proceedings of the Eighth International Conference on Creationism (2018)

but also in the numbers of total crossmatches. In terms of details, the highest ranking series (FIL1402) enjoys 52 crossmatches at P2YrsL t>6.0 out of a total of 71 crossmatches. The series located at the first quartile (the 76 th one: FIL8855) has 23 of 53 total. The second-quartile (median) series (the 153 rd one: FIL5303) shows 15 of 53 total. The third-quartile series (the 229 th one: FIL8880) is down to 9 of 40 total. Finally, the last 16 series have only 3 or fewer t>6.0 crossmatches per series. As for the P2YrsL t>7.0 in this 305-member subcollection, these decline even more steeply: 29/71, 14/53, 9/53, and 4/40. In the last 16 of the total of 305 “picked up” series, there are collectively only 13 total t>7.0 crossmatches. Now, it is normal for P2Aut to first “pick up” series that exhibit superior t-value pair-crossmatching characteristics, but, sooner or later, P2Aut runs out of them, and has to “settle” for series that are inferior in this respect. Using this conceptualization, one can realize that P2Aut has to “settle” for inferior series, in the case of TRN and FIN, a lot more than it does for many extant-tree Scots pine chronologies. This, in turn, means that both the “cluster” and “bridging” processes, of the Disturbance-Clustering hypothesis, are relieved of the burden of producing or involving series that generally have high pair-crossmatching characteristics with each other. I now consider the relative “strength” of the 5634 BC-1000 BC TRN/FIN chronology interval. This is based on both the additive (Part 2, below) as well as the interactive (Part 3) aspects of tree- ring chronology-building. 2. Interlocking Strongly-Crossmatching Paired Series: Implications of Their Uneven Distribution The total number of trees crossmatching with each other, at a given point in the chronology (sample depth), may not be as significant as the number of series, at that point, which pair-crossmatch strongly with each other. Thus, for example, a sample depth of 5, where all the series reciprocally pair-crossmatch strongly with each other, may actually be more robust than a sample depth of 10, where few if any of the series reciprocally pair-crossmatch strongly with each other. [In Figure 2, the reader sees the illustration in terms of a local set of series that match weakly with the master chronology (which is discussed in the next section). However, the same concept can be extended to a local set of series that pair-crossmatch weakly with each other (which is the subject of this section).] To further illustrate the latter, the reader is asked to imagine a floating island held together by the entwined roots of plants. It turns out that only a relatively few plants have many long roots, while most of the plants have few or no long roots. So only a relatively few plants do all the work: Most plants contribute little or nothing to the cohesiveness to the island. Moreover, what is the most important is not the total number of plants per unit area of the island, but the even-ness of the distribution of the relatively- uncommon many-long-rooted plants. Therefore, areal zones having few or no many-long-rooted plants are zones of weakness at which the island is likely to break apart. In this analogy, the island is the long chronology; the long roots are the strong individual- pair crossmatches (P2Yrsl t>6.0); the many-long-rooted plants are the “many6matchers”, the plants having few if any long roots are the “few6matchers”. Clearly, the local abundance of reciprocally- crossmatching series (sample depth) is not as important as the local abundance of “many6matchers”. Consequently, “links” in the chronology which consist only of “few6matchers” are relatively weak ones at which the chronology can more readily be breached. My survey of the pre-1000 BC part of TRN/FIN identifies a huge range of “many6matchers” and “few6matchers”. Some series are endowed with 50-60 pair-crossmatches, with the available overlapping series, at P2YrsL t>6.0, while, at the other extreme, quite a few have few or none. In order to determine if aggregations of “few6matchers” cause local weaknesses in the chronology (Figure 7, top), I successively removed all the “few6matchers” having 0, then 1, then 2…then finally 8 pair-crossmatches at P2YrsL t>6.0, and noted whenever the TRN/FIN sequence of overlapping series was breached (narrow ovals, top of Figure 7). Recall that, when submaster chronologies are being compared with each other, a minimum OVL of 70 years is required before the sufficiently-high-t crossmatch is considered credible (Larsson and Larsson 2018). In this exercise, I was more lenient: I reckoned the chronology breached whenever the last-standing local two series’ crossmatching OVL fell to <=60 years. (However, many of the breaches were absolute, that is, no OVL remaining). The results (Figure 7, top) show that the chronology becomes breached, at just the removal of the worst “few6matchers” (those with zero), and more “holes” appear as “few6matchers” with up to 8 pairwise P2YrsL t>6.0 crossmatches are removed. The 2 nd , 5 th , and 6 th millennia BC are the most susceptible to “holing” by the removal of few6matchers. What’s more, the schematic nature of Figure 7, top, does not tell the full story. The removal of successive “few6matchers” (0, 1, 2,…8) not only introduces “holes” at new locations in the chronology, as is shown: It also frequently expands the previously-made “holes”, often considerably. In fact, some of the “holes” in the 5 th - and 6 th millennia BC grow into multi-century “chasms” as the (0, 1, 2…8) removal process proceeds. However, the severity of these “chasms” is tentative, as this early part of the chronologywill be reinforced, and at least potentially extended back in time, through the soon-to-be addition of new series, according to a Finnish dendrochronologist (personal communication). 3. The Iterative “Peeling Away” of the Constituents of the TRN/FIN Master Chronology\ How well do the TRN/FIN series “bond” to the common variance? To help answer this question, I describe, in this section, my experimental removal of the series that crossmatch relativelyweakly with the master chronology (Figure 7, bottom). I variously define “weak” as t<=10, 11, 12, and 13. This definition of “weakness” is based on the fact that master chronology effect values in the 10- 13 range are commonly achieved even by small but false master chronologies (Table 2). In order to proceed, we need first to consider how TRN and FIN series crossmatch not as pairs, but as individuals against their respective master chronologies. [I did not perform this experiment forTRN/FIN in combination, because of the difficulty of transferring so many results to Notepad and then Excel]. I constructed both master chronologies, and then employed the “Test Towards Rest of Collection” function of CDendro to show the t-value of each series against its master chronology (without itself). Here is the results: TRN series, against the TRN master chronology, show a range of Woodmorappe ◀ Tree-ring chronology shortening via disturbances ▶ 2018 ICC 666