Perspectives on Information Stability in the Galactic Library During the Last 227 Hab-Years: Considerations of Scale and Scope.

Presented by Syrtainq Es a-Synthian at the ninth meeting of the Inter-Clan Conference of Alliance for Progress Associations for Encyclopedia Development, Library Studies, and Synthetic Research.(1)

As every sapient knows, it is regarded as fact that the data content of the Galactic Library is essentially static.  This is largely due to the great antiquity of the Library --elements of the database date back more than 231 (that is, 2.15*109) hab-years(2).  Given the length of time involved, if the amount of data in the Galactic Library were to double over the next 231 hab-years, then over the next 215 (i.e. 33,00010) hab-years the amount of information in the Library will increase by 2-16 (or 1.53*10-5) of the current amount.  Even assuming that the amount of data in the Library will increase by a factor of four or eight over the next 231 hab-years, the change in the amount of data over the next 215 hab-years would still only be between 2-15 (or 3.05*10-5) and 2-14 (or 6.10*10-5) of the current amount.

In fact, most Galactics would seriously question the idea of linear increase in the amount of Library data, and would regard a geometric increase as nearly impossible.  It must be admitted that once one removes data on "new" life-forms, updates to stellar charts, ongoing commercial and economic data, historical records, records of art, and so on --the amount of significant new O-2 accessible data added to the Library over the last 228 hab-years (27010 mega-hab-years) is only about one-sixteenth what one would expect if the Library's data increased linearly. (See relevant bibliography.)

Notice that this figure is consistent with the widespread belief that very nearly all information knowable to Main Sequence sapients --or at least the vast majority of durable information worth knowing --was discovered by those races that lived previously(3).  If one accepts the theory of finite utilitarian knowledge, then there exists a second reason for the stability of the Library's database.  The database is stable mainly because little remains that needs to be added --except for information of an ephemeral or historical character.  In addition, any new meta-data or durable knowledge that might be added to the database will be discovered at ever decreasing rates due to the increasing competitive advantage of Library research (in conjunction with collation and synthesis) over ever-decreasing relative marginal returns to direct inquiry.

There are other reasons many Galactics object to the search for novelty.  In contrast to those clans that pursue novelty for aesthetic reasons despite relatively low marginal utilitarian returns, even more Galactic clans view the pursuit of novelty as "tasteless."  These sentients believe that pursuit of novelty is (statistically) linked to vanity, individualism, and extreme gaucherie.  Furthermore, seeking newness is often viewed as inherently disrespectful of the Ancestors, and certainly un-dutiful.  It would be better to spend resources on those matters of greatest importance to the Ancestors --such as uplift, the Library, or maintaining ecologies.  Worse, the quest for novelty can lead to dangerous flights of hubris.  It is not unheard of for Main Sequence sapients to rediscover technology that our wise Ancestors had hidden for our own good --the technology itself being inherently dangerous to Civilization.  The quest for novelty can also lead a race to overly rapid senescence and Retirement.  Even fairly minor discoveries can disrupt the status quo and promote anarchy.  Indeed, some believe that the pursuit of novelty reveals contempt for the established galactic order and the Traditions and a desire to rebel against the wise Traditions of the Ancestors(4).  Thus (according to those with this belief), the pursuit of novelty is, in the final analysis, seditious --tantamount to a crime against Sapience and Sentience. (See relevant bibliography.)

We remind the reader of the technical distinction between "novelty" and "progress" --where novelty involves the discovery of new knowledge or the production of a truly new technology, and progress simply refers to efficiency.  As such, progress can result from new applications or combinations of old knowledge, and statistically the majority of progress--as documented by the Library (see relevant bibliography) --seems to consist of the more mundane sorts of progress.  It is thus understandable --if ironic(5) --that there is less negative Galactic sentiment toward progress than toward novelty.  Also, progress is of more obvious use than novelty, so there are few marginal-utility arguments against progress --at least in the immediate sense.  The most common argument against actively seeking progress is that progress --like novelty --undermines the stability of the Galactic order.  Since anarchy and chaos are the primary enemies of sapiency --at least of civilized sapiency --and since there are innumerable historical cases demonstrating that sapiency in a state of barbarism has a strong probability of producing its own extinction, and since progress --per se --is NOT necessary; the greater good of sapiency demands curtailment of the quest for progress.  (See relevant bibliography.)

Far be it from the Synthian race to add to the acrimony of this debate on the desirability of changing the Galactic order or the inevitability of civilizational dynamism.  Our official position has been, and remains, that this debate itself is quite dangerous to the peace and prosperity of O-2 civilization.  Nevertheless, all concerned --and certainly Librarians and scholars of the Library in the Alliance for Progress member races --see the utility (and the beauty) of critically examining the law of data stability in light of actual social behavior and data consumption. To this end, our research group at the Synthian Bureau for the Development of Alternative Information Storage and Retrieval Systems studied the marginal change in data stored in nano-scale archives distributed by the Tanith Regional Branch Library between (roughly) 227 hab-years ago (13410 Mya) and 224 hab-years ago (1710 Mya), or a span of (4+2+1)*224 hab-years (11710 Mya)(6).

For each 102410 hab-years sampled, we examined the automated standard Library subject indexes and cross reference files by machine for 25610 libraries randomly selected from all recorded 2-64, 2-48, 2-32 units issued by the Tanith 2-3 Sector Branch Library.  The total sample thus consisted of 7*222 (roughly 29 000 00010) of the smallest standard Library units.  We also machine sorted each unit by function-category based on original Tanith Office of Data Extension records, ultimately grouping the nano-branches into 3710 standard Library management function-categories (military, educational, administrative, agricultural, uplift, commercial, technical, and so on) and numerous sub-categories.  Then, using well established time-series analysis and statistical noise-filtration specialist-system algorithms, we examined subject and cross-reference category heading continuity over time.  The findings discussed in this paper are based on statistics that exclude inherently ephemeral subject categories --such as species active in the Galaxies or (contemporary) current events, as well as "accidental" entries likely due primarily to subscriber peculiarities.  The team was also interested in inter-unit variability within the broad function-categories --controlling for time.  (See relevant bibliography. Technical data and the original database available upon request from the presenter.)

Analyzing variation in the data showed that in any function-category a large part of the non-ephemeral data was indeed resistant to change.  Some function-categories, (for example, pedagogy and computing) were quite resistant to change.  The six function categories with the lowest margins of change in non-ephemeral _and _non-accidental content changed by between 2 per_128 to 6 per_128 over the 7*224 hab-years covered by the study.  Given the law of data stability, these lowest rates of change were somewhat higher than expected --but still arguably slow --by predominant O-2 standards.

However, in a few function-categories --most notable the "technology" and "military" function-categories --there was considerable marginal change in our sample of very small, hence highly selective, libraries.  For the technology function-category, non-ephemeral _and _non-accidental data changed by approximately 28 per_128 during 7*224 hab-years.  However, the military category approached a delta of 64 per_128 over the 7*224 hab-years.

Of course, not all variation in non-ephemeral _and _non-accidental change in nano-branch data over time can be clearly attributed to progress or novelty per se.  After analyzing the variation in the subject topics we are 127 per_128 certain that 35 per_128 of the variation over time in subject categories can be attributed to progress or novelty(7).  Of the remaining 93 per_128 variation over time in subject categories, fully 47 per_128 can be clearly attributed to other factors such as fad, style, or random drift.  The remaining 46 per_128 of variation in non-ephemeral _and _non-accidental variation could not be clearly correlated with any factor.  Note that on one level our findings support the contention that novelty is not a very productive field of pursuit with only 2+2-2 per_128 of all non-ephemeral _and _non-accidental variation over time being attributed to novel subject categories.  The above findings are quite consistent across all function-categories.

We were --however --gratified to learn that fully one-in-three of all progressive entries (12 per_128 of all non-ephemeral_and _non-accidental variation over time) cited at least one novel entry in their chain of synthesis.(8)

In addition to the main study using the 7*224 hab-year sample, our research group has done a parallel study using records volunteered by Clan Synthian and other friendly polities since 224hab-years ago (see footnote six).  From this data it appears that the Galaxies have been in a generally innovative phase since the spate of ecological disasters of (4+2+0)*214hab-years to (2+1)*214hab-years ago.  This is not surprising, since over the medium term rates of progress vary on an occasional, non-periodic basis, and times of more rapid progress often correlate with increased political instability in O-2 civilization.

We also examined the last 204810 hab-years intensively, using the volunteered sample.  Since the Terran-Tymbrimi contact there has been a pronounced spike in the accession rate for novel subject categories. This is does not appear to be entirely attributable to Clan Terragen and its A4P allies.  There are indications that discovery of the Humans and their clients has sparked a minor fashion for novelty, in addition to the more visible reaction against novelty and progress.  Results of preliminary data analysis are also consistent with the predicted lagging synergistic effect of novelty on the overall rate of progress.

In conclusion, our findings are not inconsistent with the law of data stability.  Considering the 20 database of the Great Library over the long term, non-ephemeral information available to Main-Sequence sapients is now increasing very slowly and the proportional rate of increase is itself decreasing.

However, as the Terragen proverb says --over the long term all are dead.  In absolute terms, the relative amount of progress in Galactic Society is decreasing rather slowly.  Furthermore, judging by the prevalence of progressive entries in nano-branch libraries with technological and military functions, progress probably confers considerable marginal advantage in highly competitive fields.  While the law of data stability is thus generally valid, in specific situations one must consider the importance of the marginal utility of progress (and of marginal change) for the topic in question.  We also conclude that there is a moderately strong correlation between wide-spread inter-clan political instability and the rate of progress.

We further conclude that ongoing progressive synthetic research is important, and a valuable contribution to the future heritage of all sapients.  Furthermore, progress is valuable in many fields of endeavor, and can be a crucial element in a clan's economic and military preparedness.  In contrast, novelty alone is of limited utility --and is only a tiny fraction of all progress.  However, novelty synergistically promotes progress when incorporated into synthetic research.

1. Syrtainq Es is a research fellow in the Synthian Bureau for the Development of Alternative Information Storage and Retrieval Systems.
        Funding was provided by The Galactic Library Trust for the Social and Historical Study of the Great Library and by The Synthian All-Clan Progress Trust.
        On behalf of the research team Syrtainq Es would like to express his sincere thanks to the staff of the Tanith Regional Branch Library's Office of Data Extension for their generous assistance and enthusiastic participation in this project.
        Translation of the original paper by Intergalactic Communication Associates under contract to the Terragen Confederation Information Agency Department of Information Services.

Return to text

2. I [T.C.S.] find the dura system quite cumbersome.  I prefer to use base two instead of base six systems (this would be a very common option in Galactic civilization.)  Also (mostly for ease of play --I want to run an Uplift campaign remember) I have decided that tradition (by extremely convenient coincidence) puts the mean distance of the Progenitors' homeworld almost exactly eight light-minutes from the surface of its sun.  This unit --called a hab --thus becomes an alternative unit of time (and distance) in O-2 civilization.  (The hab is inconveniently short for most H-2 species.)  It follows that the natural unit of 2-3hab is almost exactly a minute long,  23hab is about 64 minutes long (conveniently close to an hour), and 216hab (65536) is 524288 minutes --or a little over 364 days.  Thus 216hab is commonly called a hab-year in Anglic.

Return to text

3. The dogma of the Library data's stability makes allowances for "current events" of an historical nature --such as wars or the uplift of a new species. (However, even this kind of "ephemeral" data has little impact on the contents of the Library during the life-spans of most O-2 sapients.)

Return to text

4. The reader should note that I [T.C.S.] tend to think of prevailing Galactic values as having much in common with Confucian morality.  Both emphasize respect for the ancestors, filial loyalty and piety, and both fear of the collapse of civilized order much more than the costs of a BAD civilized order.

Return to text

5. I [T.C.S.] realize that most canonical and deutero-canonical sources say Synthians lack a sense of humor, as do most O-2 species in the Five (now four) Galaxies. However, this essay--in large measure --has been written as background for an Uplift role-playing campaign that I hope to game master.  For my quantum variation on the Uplift Universe I have made the following stipulation: There exist species with humor (these are rather rare), species without humor but having irony (these are not uncommon), and utterly humorless species (the plurality of O-2 species).

All species from Earth of Mammalian or Avian background will have senses of humor, unless it is suppressed during uplift.

Thus Humans, Neo-Chimps, Neo-Dolphins, Neo-Gorillas, etc. all have humor.  Wazoon ab-Synthian also have humor.  Tymbrimi and their clan have humor, as do the Kanten.

Soro (and most of their clients) and Synthians have irony.  (Traki also have irony)

Thenannin, Johpur, Gubru, and Tandu are humorless. (Though I am considering reclassifying Johpur as capable of irony.)

Return to text

6. Access to the records of smaller archives issued after 224 hab-YA were unavailable due to the Great Library's confidentiality policies.
        However, our research group has done a parallel study using records volunteered by Clan Synthian and other friendly polities. This sample is far less representative than the much more rigorous 7*224 hab-year sample, but cross-sample comparisons indicate that bias in the newer sample is not critical except for arguments where small changes in levels of statistical significance become crucial.

Return to text

7. Our operational definition of "progress" included those subject categories or sub-categories not entered in the Tanith branch more than 213hab-years (that is, 8192 hab-years) prior to the issuance of the sampled nano-branch.  Our definition is broader than the Great Library's definition of progress (adjusted for time).  Of those subject categories we classified as progressive, approximately 20 per_128 were not labeled as progressive by the Library.   Subject categories listed as progressive by the Library that were not included in our count of progressive subject categories numbered about 5 for every 128 subject categories on our list of progressive subject categories.
        We classified "new subject-categories" as those flagged "novel" by the Library that also occurred in our list of progressive subject categories.  The Library considers entries novel if they are not directly or synthetically wholly derived from earlier entries and are younger than 215hab-years.  Thus, our definition of novelty was more rigorous than that used by the Library, counting rather more than one subject as new for every four listed as new by the Library.

Return to text

8. This finding was also consistent across all function-categories of nano-branches, except for technical and military function nano-branches. Technical and military class nano-branches used novel elements in progressive entries at rates slightly in excess of 64 per_128, and were excluded from the above figures as outliers.
        The calculated rates of synthetically incorporated novelty would be higher using standard Library definitions of novelty.

Return to text

Return to G.I.P.P.N. entry

Return to Main Index