Read full article - Virtual Maastricht McLuhan Institute

jumentousmanlyInternet και Εφαρμογές Web

21 Οκτ 2013 (πριν από 3 χρόνια και 8 μήνες)

90 εμφανίσεις


1


Kim H. Veltman

Towards a
M
eaningful Web of Knowledge
:

"
Computer Engineering and Innovations in
Education for Virtual Learning Environments, Intelligent Systems and Communicability:
Multimedia Mobile Technologies, Experiences in Research and Quality
Educational Trends
"
(Informatics and Emerging Excellence in Education collection).
Brescia:
Blue Herons
Editions. Series Volume: I
, 2013
. ISBN: 978
-
88
-
96471
-
14
-
2. DOI: 10.978.8896471/142. In
P
ress.


…………………………………………………………………………………………………

Abstract

Initial
vi
sions of the Internet were about complete access to all knowledge. Thus far, these
visions have been hampered by three forms of compromises: technological, conceptual and an
object focus. There are also implicit contradictions in the ways we organize and s
earch for
information and knowledge on the Web. We want to find something particular and yet we use
single words, which are universal.
The semantic web

entail
s

only subsumptive relations: what
and who. Needed is a fuller approach that treats who as living
entities
,

separate from what,
and includes

determinative and ordinal relations

which are
basic
aspects of human life and
knowledge
: where, when, how, and why
.

This paper outlines a new approach to
linking
knowledge in four stages
: 1) connecting
letters,
wo
rds and
terms

with their particulars: attributes and relations
; 2
-
3
) linking these with
the
ir
sources

and with
alternative
sources
, 4) linking these with questions
such that personal (who),
geo
-

(where), temporal (when), conditional (how) and causal (why)
subsets can more readily
be
found.
It suggests linkology as a new tool in determining the veracity of claims and points
to a new Knowledge Coding Classification (KCC).


…………………………………………………………………………………………………
.


1.

Introduction

Initial visions of the Int
ernet were about complete access to all knowledge.
Thus far, these
visions have been hampered by three forms of compromises
: technological, conceptual and
an
object focus.

First, e
arly technological compromises brought limitations to this vision. The
World

Wide Web (WWW) revived the initial vision with a quest of theoretically linking
anything with everything.
Then
conceptual compromises
again brought limitations to this
vision, by
focussing on

the born digital realm
,

and
through
a particular definition of
semantics.
Third, a
n emerging quest for an Internet of Things,
is
introduc
ing

new
compromises in its fixation on things (objects).

There are also
implicit contradiction
s

in
the ways we
organize and
se
arch

for information and
k
nowledge

on the Web
. We want to find something particular

a
nd yet we use single words,
which are universal.
Linking is a key. Linking truples is
ins
ufficient because
these
entail only

2


subsumptive relations: what things are, isolated from determinative and
ordinal relations:
who, where, when, how and why
as

aspects of human life and knowledge.

This paper outlines a new approach to knowledge in four stages. First,
in addition to using
truples to connect universals via
is

and
has
, we need to link
letters,
word
s,
terms,
and
names

with their
particulars:
attributes and relations. Second, each of these needs to
be
linked with
their sources.
Third, because there are multiple sources, w
ith

chang
ing

opinions and claims
over time and space, the linked attributes and r
elations need to be geo
-
temporally referenced
to reflect different and even contradictory sources. Fourth, in order to make the immensity of
this information and knowledge accessible this corpus of links needs to be linked with
questions such that personal

(who), geo
-

(where), temporal (when), conditional (how) and
causal (why) subsets can more readily be
found
.


1.

Early
Internet
Compromises

The pioneers responsible for the early
Internet
1

made at least
four

fundamental compromises
in: a) choosing a bits and bytes model; b)
favouring
natural language
c) a narrow defin
i
tion of
hypertext,
and
d
) adopting limited data models.
These
decisions
were fully re
a
sonable at the
time because they provided a practical sol
ution for the challenge of rend
er
ing analog text and
images in electronic form.


1.1.
Bits and Bytes
2

In the 1930s and 1940s, when pioneers were preparing a framework for the Internet, the
challenges of
early computing
were largely in solving a technological
problem of how to
translate bit and bytes into visible letters, words and images on a screen.

The choice of a 2 bit
(on/off) system with 8 bytes
was
a vision of Shannon
3

and became
a pragmatic solution
in the
context of technological constraints

through
a
decision made by AT
&
T
.

T
his
binary
approach
entailed
a Booleian logic (either
-
or) and other limitations which inspired Norbert Wiener to
develop cybernetics.
It also removed meaning from information:
"The word information, in
this theory, is used in a spec
ial sense that must not be confused with its ordinary usage. In
particular information must not be confused with meaning."
4

Having defined information as
independent of meaning, any discussions about meaning could be dismissed as “just a matter
of semanti
cs.”
5


1.2.
Natural Language is not enough

A military context, which required precise yes/no answers favoured this binary
system

and a
natural language approach that avoided complex terminology
, thus

facilitat
ing

a

binary yes
-
no
process

for command hierarchies especially in the context of C3 (command, control,
communication)
.

A humorous illustration of s
erious limitations of natural language in
isolation
is

offered
by a recent
Exam for Seniors

(Appendix 1)
.

Here natural language in
isola
tion would not be
able to
provide a single correct answer
.
For instance, t
here is no way of
knowing from the words
in isolation
that the Hundred Years War took 116 years
, or that
George V’s first name was Albert.

If we use natural language it must be linke
d with terms and
dictionaries in order to recapture the meaning of isolated words.



3


Relations

Subsumptive


Determinative



Ordinal

Entities
-

Attributes

Activities




Dimensions

Who?
-

What?


Why?
-


How?



Where?
-
When?

bio
-


onto
-








geo
-

chrono
-

Substance




State



Place


Time




Quantity




Action


Affection

Position

Quality









Relation


Causes


Formal


Final


Efficient Material



Table 1.
Three types of relations

based on Perrault and Dahlberg
.
6

1.3.
Narrow Hypertext

The initial Internet began with a Transmission Control Protocol and an Internet Protocol
(TCP/IP). This enabled online electronic exchange of information. A next challenge lay in
c
reating links between between words and images in different passages, texts, on different
websites. Initial work began offline in the form of markup languages. Hypertext became the
fashion. Markup languages promised to provide tools for access to contents.

Generalized
Markup Langauge (GML, 1968) was followed by Standardized General Markup Language
(SGML, 1978ff.). These offered answers which, at the time and even today, are too complex
for regular user
s
: i.e. only a specialized programmer could be a linker.
7

In an effort to
simplify
the markup process, Extensible Markup Language (XML1.0, 1996)
8

was developed,
and
subsequently
became
an online version
through
the W3C. This was a great contribution but
still
remained too complex for regular users without
training in computer programming.

On the positive side, a Text Creation Partnership (TCP) in conjunction with Early English
Books Online (EEBO) and Proquest has made
25,363 texts available in XML/SGML in phase
1 with a further 45,000 full texts planned for

phase 2.
9

These require a subscription.
Meanwhile, another partnership (ECCO
-
TCP)
,

Eighteenth Century Collect
i
ons Online
, gave
access to 205,000 volumes, with 2,231 “freely available to the public.”
10

While still a
relatively small number of the complete c
orpus
,

this
entails
important new access t
o electronic
full text documents.

1.4.
Is a is not enough

In terms of specific collections of information, e
arly relational databases
also
promised
a
nother offline
solution. However, in trying to reduce everything to an “is a” phrase they
imposed great limitations on information. It may be true that John is a man but if he can only
have one “is” there is not very much much we can know about John. Multirelational data
bases
promised more again, but typically remained limited to
subsumptive relationships, with no
ability to distinguish between meronomy (part of) and holonymy (having parts).
Determinative and ordinal relations were ignored (Table 1).

2.

Conceptual
WWW
Compromises

In 1989, Tim Berners
-
Lee (CERN) published Information Management: A Proposal.
11

It was


4


initially a “practical project”
12

intended to make accessible information about people and
objects at CERN.
13

A working prototype appeared on 25 December 1990.

Very quickly the
system inspired a World Wide Web (WWW)
,

which transformed the scope of the earlier
Internet.
The vision of linking anything with everything emerged anew.

2.1.
HTTP and
HTML

The building blocks of the new vision were HyperText

Transfer Protocol (HTTP) and

HyperText Markup Language (HTML).
14

HTTP
permitted one site to be linked with another

online site
. The good news was that it worked. The less good news, as observed by visionaries

such as Ted Nelson, was that the links were uni
-
directional rather than bi
-
directional. Hence,
linking a website on the United States (US) with a website on California, provided no links
back to the US site if one were starting from the California site. Equally problematic was that
there was no inbuilt

system for dealing with changed website names, or defunct sites (broken
links).
15


While
the details of HTTP and
HTML remained too complex for everyday users, the

new
protocol and markup language

soon led to browsers that proved immensely useful
.

The
HTML
draft
(1993)
led to NCSA Spyglass, MOSAIC. HTML 2
(1995)
saw the advent of
Netscape Navigator and Internet Explorer (IE).
16

T
he immense growth of the WWW was
directly linked with these new tools. It had taken 30 years for the Internet to reach 1 million
use
rs. From 1990
-
1995
,

the WWW grew from 1

million to 50 million users. In the next 5
years
,

it grew to 200 million. By 2010, it had reached 2 billion users.

This immense growth was made possible by
pragmatic steps
in hypertext protocols and
markup
language,
which perpetuated a narrow approach. As the
u
sability glossary notes:



Some variations of hypertext that the web does not typically support include:

allowing people to link from any document, including ones that they don't own creating links o
f
different "types", such as distinguishing between "definition" links, "see also" links, and "author" or
"source" links (which would lead to information about the author or source text of a passage) allowing a
single phrase or graphic to have links to mul
tiple destinations.
17


Like the original practical project of Sir Tim Berners Lee at CERN
, the HTTP an
d

HTML
solutions address persons and objects. They focus on two (who?, what?) of the six basic
questions (cf. table 1).
A tacit assumption is that only un
iversally true information is valid as
is often assumed in mainstream science.



2.2.

Born Digital Realm

Another tacit assumption is more subtle. As its title suggests, the World Wide Web
Consortium (W3C) focusses on born digital materials. Hence, the
potential goal of linking
anything to everything is limited to linking anything already on the WWW to everything else
on the WWW. As long as the practical project was limited to CERN, using only CERN
documents
,

then this tacit assumption was fully reasonab
le. But in the context of
a
world
-
wide
web
,

this assumption imposes serious limits to the parameters for testing veracity, as will
become clear when we examine their solutions of RDF and the semantic web.


5



Table 2. Internet of Things



2.3.

RDF and
Semantic Web

With http and HTML, born digital materials emerged, as the Internet transformed into the
World Wide Web (WWW). A quest to define basic standards for describing these materials
led to the Dublin Core. Within a new World Wide Web Consortium (W3C) this quest
led to a
Resource Description Framework (RDF) and a Semantic web. The sales pitch version claimed
that the web should be able to link anything to everything.
The reality
became tuples/
truples:

An item in RDF is 3
-
tuple (Subject, Predicate, Object), and 3
-
truples connect to form a
Graph.”
18

In traditional grammar
,

a subject (noun) has a predicate,
c
onsisting of a verb and an
object. In RDF, the predicate is reduced to a verb and the object is separate

(cf. Appendix 1)
.

The creators of the RDF aimed to create an objective, value
-
free model. The framework
(RDF) tests the accuracy of the truple construction rather than the contents. Hence, the two
statements “John went for a walk” and “John walked 200 miles in 2 minutes” ar
e equally
correct in terms of truple logic, although the latter, in terms of logic and the human condition,
is obviously false.
19

The good news is that RDF is non
-
judgemental. The bad news is that it
has no means of determining the veracity of a statement.
RDF tests the correctness of
construction of truple statements, but has no means for testing the truth of statements that
have been constructed. The
current
semantic web is effectively a syntactic web
or imperfect
grammar web
. It is about structure of stat
ements,
not
meanings of terms.

2.4. Ontology and Ontologies

These assumptions have curious philosophical implications. In Ancient Greece, ontology
entailed the nature of being and reality.
20

So there was theoretically only one ontology. In the

6


W3C approac
h the abstract
, universal,

RDF framework is “reality”

and specific
, particular


meanings become ontologies.


Object Compromises

Tim Berners Lee’s original paper on Information Management

(1989)

at CERN was prophetic
in recognizing that the practical project was excellent for trying new object
-
oriented
programming techniques.
21

A decade later, Kevin Ashton outlined a vision of an Internet of
Things (1999),

moving from “Radio Frequency Identity (RF
ID) Tags for facilitating routing,
i
nventorying and loss prevention


in 2000 to a Physical
-

World Web with “Teleoperation and
telepresence:
a
bility to monitor and control distant objects


22

in 2020

(Table 2)
.


This is a great step forward from a web of b
orn digital objects
(WWW)
to a web that extends
to and includes a Physical World Web

(PWW)
. At the same time
,

it suffers from precisely the
same compromises and shortcomings of the original Shannon and Weaver information model
in which both humans and obje
cts are reduced to entities

and

treated identically.
There is a
danger that the quest to control drones at a distance, extends to robots and then

to humans
themselves

through brain
-
computing.
To restate the problem more dramatically: a quest for an
objecti
ve model that objectifies the physical world to the point of removing the human (bio
-
)
dimension from the model, also eliminates the value of the human science
s

and humanity
itself.


This
leads
essentially
to
a now Web of the present, tending towards a fut
ure web in an
upcoming iteration.
There are scenarios for what could go wrong but no visions of how things
could improve: bad possibilities without good possibilities and playful impossibilities. The
model reflects the quadrivium, omits the trivium and eff
ectively omits
the
past
.

History,
language, literature, philosophy and religion have no real place in the narrow goals of this
web. There is a
need for historical knowledge
, a need for worlds of imagination, belief,
phantasy, dreams.

Verbal Universals

Partly because the founders of the Internet and WWW have forgotten
or ignored
the details
underlying grammar and language, other

compromises are built into our search engines
Words are about universals. We want to find particulars and yet we use single wor
ds that are
about universals.
If we type the word dog, we are searching not only for my dog Rover or my
sister’s dog Fifi. We are potentially searching for all dogs real and fictitious
,

which have ever
existed.
In practice, t
yping

dog

in
Google brings
1,41
0,000,000 hits. Typing
Dog Rover

narrows the search down to
57,800,000 hits
,
23

still rather a lot even if we have all day.
24

This same problem applies to common names
.

Hence, Jo
hn

Smith
generates

1,170,000,000
results
,
25

while John Doe gives 52,900,000 hits.
Unless we can narrow searches with geo
-
temporal and other parameters such immense lists remain virtually useless.
Underlying these
problems are basic differences between the natural and humane sciences
.

3.

Science versus Human Sciences



7


Modern s
cience begins

from a premise that any claim is linked with an experiment, which has

been reviewed by peers and which can theoretically be retested at a later date. If a scientist is

sceptical of Galileo’s experiments 400 years ago, they can follow the instructions and
achieve

the same results, else

the claim is shifted to the unscientific category. Hence, a different time
and a different place (where and when) are theoretically irrelevant.
A tacit assumption in the
truples logic is that data and information are
also
about unchanging universals
.
Plato would
have been pleased.

In both pure science and RDF, all that counts is what and who

as

entities.

In the humanities and social sciences, information and knowledge are about changing

particulars
,

which entail multiple qu
estions: Who? What? When? Where? How? Why? Here,
truples logic is not enough: sextruples logic is required

and o
ften even more truples are
desirable. Experiments can be repeated. Unique experiences, by definition, cannot be repeated
exactly.
A source may c
laim that Caesar was killed on 15 March, 44 B.C. (the Ides of March).
But if we doubt the claim, there is no way of returning and saying: just checking. The source
could still be wrong, but unless we have a better source there is no way of knowing. So the
need to link to sources is an essential aspect of the process. RDF is about born
-
digital links
within the WWW. The humane sciences entail entities a) in the physical world, some of
which are only known via second
-
hand descriptions: e.g. accounts of former
cities that are
now ruins or lost;
b
) in the mental world (e.g. the characters in a story)
, c) in the spiritual
world (e.g. visions of God, deities, angels, avatars); d) in the digital world (e.g. personal
websites).



Technological optimists feel that
tagging alone will lead progressively to an augmented
mind.
26

This can lead to a social web, perhaps a web of opinions, but it can equally lead to a
tangled web of targeted advertising, propaganda, persuasion, indoctrination and even a web of
lies. Hence, w
hile the ability to link anything to everything is splendid in theory, a quest for
knowledge and truth points to strategies whereby some links are better than others
, namely,
those which take us back to the sources of the claim.

Indeed, the range of links
may offer a
criterion for the reliability and veracity of claims.

4.

New Organi
s
ation of Information and Knowledge


The web thus far is about data and information: i.e
.

it is primarily about individual entities
(what and who). Knowledge entails descriptions of and claims about entities (including
where, when, how, why).
Information is about 2 questions. Knowledge is about 6 questions.
To go beyond the compromises of th
e Internet, WWW

and P
WW
,

a new approach to
information and knowledge organisation is needed.
It must i) include letters, words and terms;
ii) link attributes and relations; iii) link sources; iv) link alternative sources; v) link these with
qu
es
tions for e
asier retrieval.

4
.1.1.

Letters, Words, Terms

At an electronic level, the Internet entails linking of individual letters at the data link layer
27

or link layer:

bit
-
by
-
bit or symbol
-
by
-
symbol delivery
.
28

At the application layer, linking is
primarily at th
e level of words.
Needed is a linking of individual letters
at the application
layer
which
is

mapped to equivalents in other languages
,

such that A in English, Alpha in


8



Library of Congress



146

GBV

elementary particle



535

elementarteilchen



5
5
4

partikel




1538

teilchenproduktion



125


teilchenerzeugung


130

houellebecqs


27

elementarteilchenphysik


1094

e
lementarteilchentheorie


8661

teilchenphysik



832

teilchenemission




25


Google




3,630,000

Yahoo




850,000

Bing





855.000

Yandex




1000

Baidu




347,000


Scietation






1,691,352

Table 3. German terms relating to Elementary Particle (Elementarteilchen) and publications in
GBV

and some search engines
.

Greek, Aleph in Hebrew, Alif in Arabic are lin
ked, as are the various meanings that have been
attributed to them. The same letter has a different position in various alphabets. For instance,
letter G is letter 3 in Hebrew and Greek, letter 4 in Sumerian and letter 7 in English and many
European alphab
ets. A complete mapping with spatio
-
temporal coordinates would help in
tracing the history of alphabet structures. This linking of individual letters at the application
layer needs to be mapped to the codes for individual letters at the data link layer.

If we entered the letter Psi (
Ψ
), we would be led to letter 23 of Greek alphabet, to the trident,

the trisula and under Why to its symbolism as Spirit of the World

(Spiritus Mundi), the light

embodied in Zeus, the righteous, judgment and a gematria of 700. This would be linked
the
sources of the claims,
with

related tamgas, symbols and signs. Letters of the alphabet would

remain analog equivalents of

bits, with the
difference that now every letter bit counts and
every bit has meaning, linked
with a larger context. These new dictionaries of signs and letters
would complement and link with dictionaries of words, with encyclopedias, and more detailed
articles and monogr
aphs.



4
.1.2 Levels of Knowledge

In everyday practice on the Web, we already make some distinctions in levels of knowledge.
If we want a quick definition for a term such as elementary particle we can type the term plus
answers.com or simply write
el
ementary particle meaning

in Google to arrive at basic
dictionary definitions. If we want more information typing
elementary particle

wiki
will
provide the

equivalent of a basic encyclopaedia article.

The wiki article includes 9 General
readers (sic) and 5 Textbooks. These offer a useful introduction but hardly constitute a
comprehensive overview.
In future
,

these different levels of information could be accessed by
means of templates (e.g. table 7).

On

the subject of physics, the Library of Congress lists 6 books on physics terminology, 85


9



physics dictionaries and 8 physics encyclopaedias. Typing
elementary particle

in the Library
of Congress Catalog

gives 146 titles. Typing
elementary particle

in the
Gemeinsamer
Verbundkatalog

(an online German equivalent to a national catalogue) yields 535 titles.
Typing the same term in German as
Elementarteilchen
yields 554 titles and also offers a
series of related terms which lead to over 11,000 items mainly in th
e form of books
(monographs, conference proceedings

(Table 3).
29


In modern physics
,

a majority of research
results
is in journal publications rather than in
books and monographs. The Library of Congress
(LC)
has other catalogues. Typing
elementary particle

in the
Commonly Used Periodicals
-

Newspaper and Current Periodical
Reading R
oom the same term leads to 4 titles.

In the
LC
E
-
Resources Online Catalog
it
produces 1 item.
Typing the same term in the Elektronische ZeitschriftenBibliothek
(Göttingen) yields
2

results,
30

while the same term in
German
yields
0

results.
31

S
imply typing the term into
standard search engines leads to results from 1000 to 3,630,000

(Table 3)
, alas with no easy way of viewing them
with geo
-

and chrono
-

filters.
Meanwhile,
the American Institute of Physics has the Scitation Index where the same t
erm

found 15749
out of 1691352 (500 returned).”
32

This elementary example
illustrates the need for a) greater
integration of databases of resources and b)
a nee
d to distinguish between different layers of
knowledge
. Library systems currently give us organi
zed hits
,

are often excellent,
but
sometimes so narrowly filtered that they do not lead to desired results

(e.g. LC E
-
Resources)
.
33

Search engines currently give us too many hits without filters and organized
subsets.
We need a system that allows
us
to
navigate seamlessly from elementary particle to a
quick definition, a wiki entry, to entries in 85 physics dictionaries, 554 book titles in libraries,
to 1,691,352 articles, and other sources

(cf. table 7)
.
There is a still a very long way to go
before we
have comprehensive access to information and knowledge at different levels.

4.2.

Linking Attributes and Relations

The semantic web in its current form is primarily about entities

and attributes
: the what of
objects
. In the current semantic web,
John has
two arms

and
John has a dog

are merely two
cases of
predicate
s

with the same structure qua truples.
Yet, t
he first statement is essential to a
definition of John. The second is not. Whether John has 1 dog or 30 pets does not change the
essence of
J
ohn, tho
ugh feeding them may affect his timetable and pocketbook.


The verb
i
s

applies to universals.
The verb

h
as

applies to both individuals (partitive relations)
and multiples.
When we are se
a
rching for particulars we need additional parameters: subsets
of is such as
is x color
,
is x size
,
is x shape
, is
in x place
. We need subsets of is that reflect
accidents (e.g. Aristotle) and
facets (e.g. Ranganathan)
.


In
Aristotle’s
approach an entity h
as
10 ingredients
:

'Expressions which are in no way
composite signify substance, quantity, quality, relation, place, time, position, state, action, or
affection.'
34

This basic list of 10 features is

described as the 10
a
ccidents,
a
ttributes or
c
ategories of

Aristotle. While their meaning has shifted
,

they remain a basis for knowledge in
the West. These 10 Accidents
35

have been mapped
to Aristotle's 4 Causes.
In addit
i
on
,

they

10


can be linked, to 3 kinds of
r
elations

(subsumptive, determinative, ordinal), 4
Categories of
Dahlberg (
Entities
,

Attributes
,
Activities
,
Dimensions
)
and to the 6 Questions

(cf. table 1).



4
.2.1.Subsumptive Relations



A brief outline of three types of relations: subsumptive, determinative and ordinal
is

useful in
helping us to und
erstand the enormity of the challenge.
Subsumptive relations deal with
entities, which are of two kinds, persons and objects, living (bio
-
) and being (onto
-
), who and
what, names and subjects/terms/words.
Living entities (bio
-
) have free will and can make
decisions. Entities, narrowly defined (onto
-
), exist without life, free will, choice and decision
making.
Mediaeval thinkers used these distinctions for a chain of being, a principle that
remains valid eve
n if metaphysical associations have changed.




4
.2.1.1. Names

Names
entail a series of challenges in addition to obvious problems of spelling.
First there are

prob
l
ems of listing them alphabetically, even in the case of f
amous names such as Leonardo

da Vinci
.
Some libraries class this under L for Leonardo; others under V for Vinci, Leonardo
da and
even
under D as in da Vinci
. Hence clicking names for Leonardo da Vinci would
remind
us
of these and other variants.

Second, names often have many altern
ative versions.
For instance
, J
ohn has 85 variant forms

in one source
:

Anno, Ean, Eian, Eion, Euan, Evan, Ewan, Ewen, Gian, Giannes, Gianni, Giannis, Giannos, Giovanni,
Hannes, Hanno, Hans, Hanschen, Hansel, Hansl, Iain, Ian, Ioannes, Ioannis, Ivan, Ivann
, Iwan, Jack,
Jackie, Jacky, Jan, Jancsi, Janek, Janko, Janne, Janos, Jean, Heanno, Jeannot, Jehan, Jenkin, Jenkins,
Jens, Jian, Jianni, Joannes, Joao, Jock, Jocko, Johan, Johanan, Johann, Johannes, John
-
Carlo, John
-
Michael, Johnn, Johon, Johnie, Johnnie,
Johnny, John
-
Patrick, John
-
Paul, Jon, Jona, Jonnie, Jovan,
Jovanney, Jovanney, Jovanni, Jovonni, Juan, Juanito, Juwan, Sean, Seann, Shane, Shaughn, Shaun,
Shawn, Vanek, Vanko, Vanya, Yanni, Yanno and Zane
.
36


Such lists of variants can be of great use when
searching through disparate historical sources
,
which typically have one of the variants rather than the standard name.

Hence
,

variants
expand the range of sources that can be accessed.

Simultaneously they can help us identify
narrow subsets of a name.
Thi
s approach applies also to groups of persons, peoples, tribes,
clans. For instance, the
Alans are one of tribes who came from Asia, settled in the Caucasus
(Ciscaucassia), a subset of whom moved further West. Typing Alans under What would
provide a minima
l definition and a list of terms pertaining to Alans, e.g. Alan Alphabet, Alan
Language, Alan Tribes etc. Clicking Who would provide a list of names associated with the
Alans:
“Alans, Alani, Alanliao, Aorses, As, Asii, Asses, Balanjar, Barsils, Belenjers,
Burtas,
Halans, Iass, Iazyg, Ishkuza, Ishtek, Jass, Lan, Ostyak, Ovs, Rhoxolani, Steppe Alans, Yass,
Yancai….”
37

The combined list would lead to a wider range of sources.
Here again, i
ndividual
variants
can become starting points for narrower searches

leadi
ng to subsets.


Clicking Where would provide maps showing Alans
,

which can be filtered geographically and
chronologically. Clicking When would provide
dates,
timelines and history of Alans. Clicking


11








Figure 1.
Three maps of what is now Russia: a) Scythia et Serica, b. Sarmatia et Scythia, c.
Russia.
38



12


How would give methods, practices, customs of Alans. Clicking Why would lead to Alan
beliefs, religion, mythology, theories, reasons, symbolism. Items found under

all 6 of these
questions need to be linked to a source (documented and with references: to use terms from an
earlier medium). Advanced versions would provide a complete bibliography of articles, books
and (serious) websites on Alans.

A third challenge of reducing the number of hits occurs with common names such as John
Smith with
1,170,000,000 results in Google.
Typing John Smith Willoughby

1580
-
1631, in
Google
[i.e. place of birth, date of birth and death]

reduces this list to19
,1
00.
If
all
names
were provided with geo
-
, chrono
-

and profession tags then typing a name could be followed
by subsets in a given city, a specific address and a specific range of dates in order to move
from over 1 billion initial results to the single result th
at interests us.

4
.2.1.2. Terms (Subjects, Keywords, Words)

The founders of the Internet and WWW favoured the use of natural language words over
terms and keywords.
39

A systematic mapping
(insofar possible)
of subject headings in various
subject headings

and library classification systems
e.g. Library of Congress Subject Headings
(LCSH),
Répertoire d'autorité
-
matière encyclopédique et alphabétique unifié

(RAMEAU),

Schlagwortnormdatei
(SWD)
and Universal Decimal Classification (UDC)
would prove a
powerful
tool in bridging everyday usage of words with professionally defined terms, all the
more so because this would link with existing library catalogues and their contents.

Four of Aristotle’s 10 accidents relate to what and subsumptive

relations: substance, quantity,
quality and relation.
40

The scope of these accidents is a topic of discussion.
41

Their definition
also changes over time. For instance, Galileo, changes the meaning of primary qualities to
include only those which can be quan
titatively measured. Hence, the linking of names and
words in historical sources with accidents needs to be complemented with changing
definitions corresponding to the date of the sources.

Some classification systems use a variant of Aristotle’s acciden
ts called facets. Ranganathan,
for instance, identified five key facets which he summarized as PMEST: Personality, Matter,
Energy, Space, Time, which correspond to who (bio
-
), what (onto
-
), how (techno
-
, socio
-
),
where (geo
-
) and time (chrono
-
). Linking w
ords in sources to these facets would mean that
they can be
retrieved as subsets using one of the six questions.

4
.2.2. Determinative Relations

This approach also applies in the case of determinative relations, where earlier qualitative
concepts of actin
g and being acted upon have been replaced by
activities and processes and
more quantitative methods, especially in chemistry.
42


4
.2.3.
Ordinal Relations

Ordinal
relations entail space and time, where and when, geo
-

and chrono
-

dimensions. At a
simple lev
el this requires that sources and the claims in them are linked with geo
-

and chrono
-

links.
In technical terms
,

we need more than an RFID or its equivalent to identify objects

13


uniquely: they must also have geographical co
-
ordinates and a time stamp.
Since

books and
manuscripts typically have a date, these dates can also be linked with the claims made in their
contents.
Of course, some sources have no clear dates an
d

there will be some cases where
different experts offer very differ dates. In such cases
,

the source of the alternative dates needs
to be added.
The range of dates associated with a source or a claim can serve as an indicator
of the uncertainty thereof: i.e. certain knowledge entails precisely documented dates which
are undisputed, while uncer
tain knowledge does not.


4
.2.3.1. Time (chrono
-
)

The WWW and its contents are organised using the current Gregorian calendar which is an

obvious choice for the “
n
ow web.” A knowledge web which includes the past will need
conversion tools to help us with alternative calendars. A number of calendar
conversion tools
already exist as specialized applications.
43

Needed is their systematic integration within the
system su
ch that clicking on a Babylonian, Hebrew,

Islamic, Julian, Persian or other date
enables a direct conversion to contemporary equivalents
, without a need to open special

ancillary programs
to

do the conversion.



4
.2.3.2. Space (geo
-
)

The past decades have
seen great advances in making geographical information accessible

online.
Google Maps and Google Street View
have transformed our sense of the possible. The
rise of location based services

is
increasingly linking given words with relevant geo
-
co
-
ordinates:

i.e. with specific companies/buildings/shops/restaurants.




Maps with a time frame


In the current Google Images
,

typing
Sarmatia
, provides a series of maps and many items that

relate to objects found in Sarmatia. Typing
Sarmatia map

provides a majority of maps which
are in no apparent order. Typing
Sarmatia map 1600
-
1700

or
with ot
her dates provides some
detailed maps and a majority of entries which are not related to the query.
Needed is an
historical equivalent of Google Maps and S
treet View. Typing Sarmatia would then lead to a
basic map. Clicking on when and adding a date or chronological frame (e.g. 1000
-
1200)
would offer the relevant subsets. A map timeline function would allow us to follow how one
map morphs into another as we
move through the centuries: e.g. how ancient Scythia became
Sarmatia and then Russia (figure 3).

Linked with this exercise would be an integration of historical toponyms and information
from gazetteers.
Today, if we type a place name such as
Urfa
we would

be directed (as
already happens in Google) to
Şanlıurfa

(Sanli Urfa) and be presented with a list of alternative
names:
Adma
,
Antiochia on the Callirhoe
,
Ar
-
Ruhā
,
Edessa
,
Riha
,
Ur
-
hay
,
Urhai

and
Ur of
the Chaldees
.

In future
,

each of these could be in a coordinated database. Books would be omni
-
linked. A
beginner’s version would link only to a simple dictionary definition (e.g. Answers.com) and a
wiki entry. Research level would potentially provide us with
a
full history of th
e toponymn.
Any variant in a source would get us back to its modern name and, where appropriate, its


14


Who



What



How



Where



When

Why

bio
-



onto
-



techno
-
, socio
-


geo
-




chrono
-


P



M



E



S


T

Personality


Matter



Energy



Space


Time

Names



Subjects


Techniques, Methods

Places


Dates Theory

Personal Nouns


Nouns

Verbs


Adjectives Adverbs





Locative


Temporal










Adverbs


Adverbs


Table 4
. Basic questions and basic parts of grammar.


4.3.
Linking Sources

Greek, Latin, mediaeval, Renaissance and other names. If we encountered Ain Zarba we
would learn that it is now called Anavarza, was called Ananzarbus, Caeserea and Justinopolis.


Maps with a spatio
-
cultural frame

Such historical equivalents of Google map
s will require a further feature, namely competing
boundaries on maps from different countries: a problem that continues to the present day. For
instance, the Indian, Pakistani, Chinese and Nepali maps of their own countries and their
neighbours differ. Po
land’s, Russia’s and Germany’s maps have often differed considerably.
Border disputes are a visible manifestation. Standard maps typically give one set of
boundaries and give no hint of the problems. In some very sensitive areas, even acquiring a
precise
map is difficult. Needed is a system which allows us to compare the same area as
defined from both sides of borders.


In the original CERN project, sources were not an issue. All the materials were officially
linked with CERN so their authenticity and veracity could be taken for granted. A narrow
view of an Internet of Things foresees that objects all have an RFID or equi
valent tag, which
answers the problem of sources: or at least theoretically, if one can assume that the ID has not
been replaced by a virtual substitute.

A Web with historical knowledge immediately poses a series of new challenges. Even the
most ardent t
echnophile
will
accept that complete retrospective tagging of the past is
impossible. We cannot go back to Cleopatra or Alexander the great and ask them to wear their
new RFID. What is possible however is to RFID historical sources (books, manuscripts,
ins
criptions). If the sources have their unique identifiers (be they call numbers
, ISSNs

or
RFIDs), then the claims
made therein can equally be given unique identifiers and linked with
these sources. H
ence
,

if manuscript A claims that Caesar was killed on 15
March, the claim
acquires an ID which includes geo
-

(where the manuscript was written and is now) and
chrono
-

(when it was written) tags).

If this is done for all sources that make the identical claim then a new kind of timeline linked

to individual claim
s becomes possible. This can function in the manner of a

c
itation index
avant la lettre. Claims with only a handful linked sources will tend to weigh less than
identical
cl
a
ims with hundreds or thousan
d
s
of sources.





15


4.3.1.
Linking Attributes, Relation
s and Different Sources

Not all cases are this straightforward. Zoroaster, also known as Zarathusthra, is the founder of
one of the major religions of the world. Indian sources link Vasistha and Zoroaster, also called
Vasistha and Vishvamitra, or
l
egitimat
e and
i
llegitimate son of the s
u
n
. Vasishta and his
followers are brahmans, linked with the Devas, worship Indra, the moon god, Chandra
,

and
are
thus
link
ed w
ith the chandravamsa (moon race)
. Zoroaster and his followers are Magi,
linked with Asuras
, worship Surya, the sun god and thus linked with the
suryavamsa (sun
race).
44


In the West, t
he
Indian connections are often omitted and the
chronology of the historical
Zoroaster is a matter of great debate. The Parsis in India speak of a date prior to 6,
000 B.C.
,

as did Plutarch. 1,750 B.C. is the date given by some. Some Iranists tend to favour 11
th
/10
th

c.
B.C. Ammi
a
nus Marcellinus claimed 4
th

c. B.C.
45

In traditional scholarship
,

a given school
would frequently accept a given date and ignore alternative

evidence. In future,
especially
in
cases of controversy,
we need
lists of the alternative dates linked with their authors and
sources.
46

In cases where there is no single true source, then we need access to all sources
claiming to be true.



4.3.2

Linking Attri
butes, Relations, Sources, Questions

Ranganathan’s

facetted classification points to a simple, yet profound insight. While
information may be about objects in isolation, knowledge entails a range of facets (PMEST)
which apply to a range of questions and
can be mapped to basic parts of grammar (table 4).
These could be mapped with verbs and prepositions found in the Integrative Levels
Classification (ILC).
47

This implies the possibility of a semantic web in a deeper sense.

The 26 top level categories of
the ILC concern what

questions
, although a few can be aligned
d
ifferently

(appendix 4)
. In ILC
, four facets deal with subsets of what:
4 made of element, 5

with organ

as well as
8 like pattern and 9 of kind (which follow the Aristotelian accidents).
Four
facets deal with other questions: e.g.

1 at time (when), 2 in place (where)
,
3 through
process (how)

and 7 to destination (why)
.
One facet (6 from origin) is effectively a history
(when) dimension applied via the questions
. The opening facet,
0 under persp
ective
,

covers
the theme of sources.
48

In search strategies, each of these facets could be aligned to basic
questions.

A future system would begin with a term (noun) such as
car (with its equivalents: automobile
etc).

Narrower terms for car would include:
luxury car, motor car, family car, electric car etc.
The broader terms would indicate its classes (is a: e.g machine, i.e. meronomy). The narrower
function would also identify the components of the car (has a, partitive relations,
holonymy)
. The narrowe
r function would also link with
verbs pertaining to a car: e.g. to start,
to run, to move, to drive, to accelerate,

to speed, to stop.
We have dictionaries and
etymological dictionaries of words. We have usage lists of words. This new approach points
to a
future dictionary, which links nouns to a specific set of verbs. Hereby
,

the scope and
functionality of any noun, object,
verb,
will become more visible.


16


This narrower function would also offer adjectives as subsets of car: e.g. of, and
, thus giving
a lis
t of associations linked with a given word.
Clicking on other questions would give access
to other facets of cars. For instance,
W
ho would provide names of automobile inventors,
companies, manufacturers, dealers, repairers. Where would locate these. How wo
uld give
information about horsepower,

fuel consumption, acceleration rate, and performance.

More
detailed how would lead to repair manuals. When would link to
dates,
timelines and also to
historical records and past knowledge of cars.

If words are mapp
ed to terms as an expanded version of
see also
,

then one could in future
have omni
-
linked books where every word becomes an entry point
to
a new encyclopaedia of
recorded knowledge
, which reflects the principles of grammar
and
eventually all the liberal
ar
ts
.
49
.

Some versions can have quicktionary
-
like pens
,

which are wirelessly linked to the
network. Other versions can be touch screen
,

as is becoming the fashion in mobile devices.
It
will not be able to access all that has ever been done
,

but
can give
us
access to the tremendous
amounts of knowledge in our memory institutions.


4.3.3.

New Philosophy of Linking and New Role for Sources

In the original Internet, linking was possible but tacitly discouraged. There was a culture of
needing to ask permission to link with another site. The rhetoric of the World Wide Web
changed this to a vision of being able to link anything with everything (
i.e. every other thing).
The practice of the W3C was to create a
Resource Description Format (RDF) for a
semantic
web in terms of truples where only this flavour of links could be “verified” and hence be fully
approved. The tacit message was: all links are

possible, only truple links are respectable and
legitimate.
To achieve technological success
,

the pioneers of the Internet removed meaning
from information. For the same reasons, the developers of the W3C removed meaning from
semantics.

Removing meaning w
as an efficient technological solution of ensuring normalised
data transfer at the data link layer.
Now the medium is the message in a way not even
McLuhan foresaw.
The good news was that it enabled programmers to focus on the accuracy
of the transmission
process. The less good news was that it removed truth of claims

from the
equation.
The pipeline was verified but its contents effectively remained unexamined.



A meaningful semantic web requires
serious
chang
es

in

practice and
philosophy. First, the
tr
uple approach needs to be refined, in the sense of differentiated, to deal with each of the
facets and each of the accidents. The same principles of verification need to be extended to
each of these more specialized truples. Then the scope of the truples “
statement” needs to be
extended to include a link to a specific document, with a specific date and place.
I
n the initial
semantic web

a truple

had a form:
John has a dog
.
I
n

the new version
,

this truple would read:
John has a dog

according to document x (call no and/or RFID and weblink

to source
), dated x

(date
, i.e. chrono
-
link
)
, in place x
(place, with link to co
-
ordinates
, i.e. geo
-

link
).
Accordingly
the validation process go
es

beyond the
initial
(specialized) truple,
and

incl
ude
s

document
/source with

geo
-

and chrono
-

links
.

In the Internet and WWW models, the source is assumed to be another electronic item
elsewhere in the system: A simply links to B via an intermediary (link) C to create a truple. In
the new model, a claim i
n A links via
a
n intermediary to a claim in B, which then links to the

17


source of that claim

(including its name, place and time). In traditional publications, the
source is appended as a footnote or as an endnote. Or more precisely, the footnote cites the
name, title, year, publication place, publisher and page of the source but ultimately does
nothing more than point to a resource that is somewhere else. Sometimes even finding
the
actual document is a minor research expedition.

In the new approach,
as in
s
upply chain management, the s
ource

is fully integrated into the
supply chain. Following the links takes us back to the original document or at least a verified
facsimile that can be authenticated via invisible (electronic) and visible watermarks.
In
parti
cularly important cases there could be final links to the physical object using webcams
and microscopic sensors.
The source
may still be
something

outside and extraneous to the
document in which it is being quoted:
but
it is now
also
an essential
part of t
he claiming
process and can be reached at any time without requiring special research in tracking down its
location.

4.3.4.

New
K
inds of Validation

and Truth

In the new approach
,

there are
now three
kinds of v
alidation
: 1) of the pipeline a
t the data
-
layer level in the Internet model; 2) of the logic of the truples

at the application layer
; 3) of the
extensions of the truples linking back to
bibliographical
source
s in memory institutions and
physical sources (sites, monuments) in the physica
l world
. This represents an ideal case.
Traditionally there has been a whole range of writings: some scholarly, with
an
apparatus of
footnotes, bibliography, appendices, indexes etc.; some more journalistic,
others personal,
typically with no footnotes
.
50

T
his range of styles is also found in websites and should
continue.
In cases where websites are purely personal expressions, they should be completely
free to do just that, within the bounds of decency and general discretion: or not
,

if the site is
for a pr
ivate group.
For these personal sites only the first 2 kinds of validation apply.

However, in cases where an author or group lay claims to being public and official, then the
veracity of their claims must be open to scrutiny.
Here all 3 kinds of validation

apply.
Morevover, sources in memory institutions have further information connected with them: a
major publisher (Oxford, Harvard)
usually
has more weight than minor publishers; an article
in a standard journal for the field: e.g. Nature for science or Th
e Lancet for medicine
,

has
more weight than others; a peer reviewed journal has more weight than an un
-
reviewed
journal. Books and articles are further linked to citation indexes, all of which can potentially
be used as factors in weighing the value, relia
bility, seriousness of a given source.

In complex cases, the “supply chain” may lead to a source in a memory collection and then
further to a memory site (e.g. museum, archaeological site, historical monument). For instance
the author is writing abo
ut Troy and cites a standard monograph such as Schliemann
concerning some detail. The link process would then go to a copy of the Schliemann
publication and then
link back to the item in Troy under discussion possibly via a museum
where that item is now di
splayed.

In the exact sciences we expect
, i
ndeed, we assume
fac
t
s.

In the humane sciences
,

there are
many
facts
. Heads of state (kings, queens, presidents, prime ministers)
, ministers, civil

18


servants, employees
assume
their position
on a precise day and e
nd on a precise day. There
are clear records
. In the case of historical sources,
there are
also
many uncertainties.
Documents may have been bombed, decayed from lack of proper care, been stolen, or simply
misplaced. In the absence of documents, sources, no

real certainty is possible. As a result,
s
ome
cite
th
ese difficulties
to
argue for relativism
and to
claim that truth is
now
an outdated
concept.
Links offer a way to defend old
-
fashioned claims to truth.

Trying to reduce information to its smallest comp
onents leaves letters and words in isolation,
without context and with no parameters for checking their truth value, their veracity. Linking
electronic letters and words to sources and
in turn to
the sources that these describe

brings
truth back into the d
iscussion. If a claim seems questionable or provokes doubts, then there is
a way to return to the evidence on which the claims are based and come to our own
conclusions. Just because we cannot always be absolutely certain, is hardly a reason for
abandoning

the very tools we have of approaching as great a degree of certainty as is possible
under the circumstances.


4.4.

Overviews

The new approach to linking promises more than a better method for verifying claims. It
introduces a possibility of making
accessible

cumulative results of scholarship
in new ways.
Instead
of a simple claim x built the Parthenon on the Acropolis in Greece, we could
potentially have a chronological list of all the architects/artists to whom the building has been
ascribed.

Instead of loo
king at the Acropolis in isolation we could trace the location of acropolises
(acropoleis) throughout Greece and the Middle East. We could study how it relates to the
citadel tradition in the Near East; the tradition of fortified cities in Persia and Turkm
enistan;
how it relates to oppidum of middle and northern Europe; the rocca
tradition of Italy and the
so
-
called Castro culture of
the Iberian peninsula.
51



The tools with which we
search,

and the depth to which we access knowledge, will vary
tremendously
depending on needs and goals. Standing as a tourist in front of a monument in a
foreign city, a snapshot with a camera in a mobile device may suffice to access basic
information. Sitting as a scholar at home, wishing to do serious research, a minimal versi
on
would be a simple monitor. In more dramatic cases
,

there could be a main monitor linked with
a five further screens, enabling me to search for something and then view details of who,
where, when, how, why on separate screens. In some cases a wall screen

might be more suited
for videos and television documentaries.
In
other

cases
,

images on two screens may be better
suited for comparing similar or nearly identical images.


4.5.

Worlds Wide Webs

The initial WWW emphasized the global character of a new tech
nology
using
geogra
phical
image
ry
. Products such as Google Maps and
p
rojects such as the Physical World Web
(PWW)

focus on this geography in a more literal sense.

There are also first attempts at maps of the
heavens

(e.g. Google sky Map). Traditionally, there were three worlds: heaven, intermediate

19


space, earth.
52

Later systems linked 7 heavens with 7 planets. We have a GIS for the physical
world. We need a GIS for earlier cosmologies in order to understand how they sa
w the
universe.
The objects (planets, stars, deities) in these cosmologies would be linked to
databases providing us with a history of individual items.


In
Dahlberg’s ICC there are 9 areas (Appendix

5
)

which can be seen as 9 worlds. Each of
these can be

linked with prefixes: e.g.
1.Form & Structure entails the Greek phylo
-
, morpho
-
.
It also aligns with a.form in the ILC.
Hence, the ICC becomes an ordering system for the
different layers of reality. Physical layers such as energy and matter and cosmo
-
geo
can be
aligned with scales for powers of 10, such that macroscopic and microscopic levels become
further ordering, searching and navigation tools: e.g.
choosing the
power 10
-
1
2

(
1
p
icometre
)
takes us directly to
a
tomic
s
tructure
,
c
osmic
w
aves
,
d
igital
-
s
tructure
,
e
lectro
-
m
agnetic
s
tructure
,
genom
e size
,
m
o
lecular
s
tructure
,
q
uantum
s
tructure

and the
u
ncertainty principle
.

In addition to these conceptual maps of the physical world there is a long history of ma
ps of
the spiritual world:
e.g.
32 letters linked with 32 stages of initiation, enlightenment in
ascending
through
32 stages of consciousness. This points to GIS of spiritual worlds
,

which
would effectively be 3
-
D versions of the visualizations in complex
Buddhist thankas.
There is
also a history of imaginary worlds which can be recreated. Hereby
,

an initial WWW will
become Worlds Wide Webs.
Fantastic Voyage (1966) and Honey, I Shrunk the Kids

(
1
989
)
,
were science fiction films about changing to a microscopic level. In future,
entry into such
phantasy world
s

could be example
s

of navigation at micro,
neuro, nano and
pico levels.

Personal visualisations of meditation can continue, but the new methods may enable shared
vision journeys.

5.
Challenges

There are many challenges to achieve such a vision, including enormous amounts of effort
and dedicated co
-
operation in a task much
larger than any small team could hope to achieve.
There are also two specific challenges
,

which are undermining and could prevent entirely the
achievement of this goal, namely: privatisation and destruction.

5
.1.Privatisation

In the past
,

there was a clear

division between public and private in the personal sphere. At
the level of countries this became a division between a public sphere which entailed activities
for the public good (not for gain or profit), and a sphere where private companies could
operate

with a view to making profit.
Reference works, which are the tools to gain access to
knowledge in our memory institutions, clearly belonged to the sphere of the public good.
They were the products of long years of dedicated work of scholars with no view t
o making
maximal profits and publishers struggling to meet basic costs.


In the last 50 years this model has shifted. Increasingly, publishers of reference works (e.g.
Sau
r, Bowker, Di
a
log
) have been acquired by private companies where profit is a domina
nt
goal
.
For instance, Saur was acquired by Reed
-
Elsevier
, then Gale
and is now owned by De
Gruyter (Berlin).
Chadwyck Healey (Cambridge, UK)

focussed on “
content collections that

20


support research and teaching in the humanities and social sciences

,
53

was acquired by
Proquest (Ann Arbor, which began as University Microfilms).
Independent scholars or
individuals cannot subscribe to Proquest: only institutions.
Proquest now sees itself as:

a gateway to the world’s knowledge


from dissertations to gover
nmental and cultural
archives to news, in all its forms. Its role is essential to libraries and other
organizations whose missions depend on the delivery of complete, trustworthy
information.
54

Meanwhile, Proquest and Cambridge [US] Scientific Abstracts hav
e both been acquired by
Cambridge
Information Group (New York),
55

and now appear as Proquest
-
CSA.
This new
company has also acquired Bowker “
the world's leading provider of bibliographic information
management solutions

56

and the Early English Books Online
(EEBO), the full contents of
125,000 early English Books.
57

As a result,
the domain of reference materials, traditionally
part of
the public domain
,

are now owned by a private company. Ind
e
ed
,

the copyright of
Chaucer, Shakespeare,
Erasmus (in English), Mil
ton, Spenser, Pope and virtually every
English author from 1475 to 1700 is claimed by a company in New York. In a best case
scenario this is a case of Americans making money from the efforts of others. In a worst case
scenario, a new boss could theoretical
ly decide that the corpus of early English literature was
no longer accessible outside the U.S.
or that the past was no longer relevant in a new world
order.




5
.2.
Short

Term Gain

One of the positive trends of the past half century has been a consolidation of earlier efforts.
The Saur Verlag, concerned with reference works, produced numerous lists of authors. These

have been integrated into a single list of 6 million names known as
the
World Biographical
Information System (WBIS) Online
.
58

This is now owned by De Gruyter.

It requires a
subscription only available to institutions.
The cumulative efforts of individual scholars to
create tools for access to knowledge are no longer freely

accessible to individuals generally
and not even to individual scholars.

T
h
e

publisher
, De Gruyter,
has introduced a new model for libraries called
Patron Driven
Acquisition

(PDA)
.
59

The idea is “
to offer users access to all digital content, but only
charge
for actual use.


60

This admirable goal comes with an a cost tag of 1,585,000 euros per library
annually.
These prices do not give the library any permanent ownership. If
,

after a year
,

they
stop, the only new addition to a library is memory
of a lar
ge bill.
The accounting model
assumes that the use of any item, database, e
-
journal or e
-
book is worth 2.50 euros. Libraries
that choose only the databases for 345,000 euros, have an accounting model where searching
1 item in a database costs
“only”
1.25 e
uros per item.

This may sound modest. A researcher working on a bibliographical project might typically
need two minutes to consult a specific item. In a hypothetical case
,

where they worked 12
hours a day,
that would be 456 euros per day. With a 5 day week this is 2,280 euros per week.
Assuming a months holidays, a year (47 weeks of work) would amount to 107,160 euros.
Assuming that the complete package accounting prices applied then a year’s dedicated study


21


would cost 214,320 euros per person.
Even hypothetical students would have problems
paying such prices and those who could afford them would probably outsource the task to an
assistant.


When the WWW began there were initial scenarios
of
telecoms
,

which
f
oresaw charging 30
eurocents per screen view, and
p
lans by national libraries to charge for the use of their
catalogs. Energetic and adroit actions of concerned citizens derailed these horror scenarios.

While posing as a cost saving device, PDAs
are

troubl
ing because
they

undermine and even
threaten the future
of research. Aside from the problem of poor students, if each library had to
pay over a million to each publisher annually and ha
d

no systematic collection at the end, the
vision of memory institution
s with systematic and near comprehensive collections would be
finished definitively.



5
.
3
.
Destruction

An even more sinister danger comes from an unexpected quarter: new practices in war. From
earliest times wars have been associated with death and des
truction. They have also been
balanced by tacit assumptions: that killing and destruction will be held at a minimum
while
a
military front advances in its conquests. This tacit assumption was especially true in the case
of cultural content. Sample items we
re sometimes taken as part of the spoils of war, but even
so
major cultural centres such as Babylon survived 5000 years of invaders including
Alexander the Great, the Huns, Genghis Khan and Tamerlane.


In the past
d
ecades
,

there has been a fundamental shift.
The museum of Bagdad had 80 % of
its collection stolen and many pieces destroyed. The museum of
Mosul (Iraq),
61

opposite the
ancient city of Nineveh was
looted and
was
also victim of one
the
first bombs that fell on the
city in the Iraq war. Recently the “terrorists”

in Mali attacked the museum and library at
Timbuktu, a precious centre
for North African manuscripts.

Destruction of libraries, mosques,
archives are becoming ever more part of a trend. In the name of fightin
g and killing an enemy,
the collective memory of some peoples is being consciously destroyed.
62

The so
-
called Arab
Spring is destroying heritage in all the countries affected.
The rhetoric is eliminating
terrorists: the practice is an increasingly systemati
c attack on the cradles of civilization: Iraq
(Babylonia, Mesopotamia);
Libya and Tunis (Punic and Carthaginian culture),
Afghanistan
and Pakistan (Indus Valley), Egypt, today Syria (Aramaic culture) and according to some,
tomorrow, Iran (Persia and Assyri
a)
.
If the sources of a people’s memory
are

removed, the
way is open for other
s

to rewrite their history.



6
.

Knowledge
, Information and Data

Traditionall
y, there was a spectrum from facts and information to knowledge and wisdom. In
India, there were parables of being too fixated
on knowledge. Hence
,

the all
-
knowing but
unwise god, Ravana, depicted as having 10 heads, was ultimately defeated by truth and
wisdom.
63

In Antiquity, the organization of knowledge was initially a task of philosophy.
From the Renaissance to the 19
th

century, it became increasingly a domain of librarians and
then library science.


22


In the first decades of the 20th century, a vision
emerged of global access to knowledge in
terms of a world brain (Gehirn der Welt).
With Otlet and Lafontaine, t
his led in practical
terms
to
Universal Decimal Classification (UDC
, 1904
-
1907
)
, to the Mundanaeum (1910,
Brussels, now Mons), and to

publication
s: Traité de documentation (1934)
64

and Monde:
essaie d'universalisme (1935)
65

in which they outlined a vision of world
-
wide network of
knowledge
.

That same year Bliss (1935) published his classification system.
A
decade
later
Vannevar Bush
(1945)
introduced

his MEMEX idea

and narrowed the vision
.

This was two
years after
Ranganathan published his Colon Classification (1933),
and
Eckert at
Columbia
experiment
ed

with astronomical data
: l
ater called the first use of “
automatic computing
machines for research work.”
66

Th
e advent of electronic
media brought changes to the spectrum

of knowledge
.
In a first
strand,
Shannon and Weaver, in their Information Theory (1948), changed the name of facts
to data,
added two items at

the lower end and also removed the final two items, such that the
new spectrum was

now
: bits, bytes, data, information.
The initial
American
pioneers

ignored
movements towards collective intelligence and a world brain in Europe and
shifted attention
to

ho
w bits and bytes could combine to store and transmit data.


This new spectrum also led to rifts
. Multiple strands developed in parallel unaware of or
consciously ignoring each other.
A
second

comput
ing

strand, in the vision of Doug Engelbart
(
e.g.
196
3 f
f.
) was fascinated by potentials

for collaborative work
“to augment the human
intellect”
67

and
to
augment Society’s collective IQ.
68

Th
is

vision
led to the mouse,
69

included
Dynamic Knowledge Repositories (DKRs)
70

and Open HyperTools
,
71

led to an
Open
Hyperdocument System (OHS) and HyperScope.
72

It led to Computer Supported
Collaborative Work (CSCW) and Computer Supported Collaborative Learning (CSCL).
A
third

computing strand by
Engelbart’s contemporary and friend, Ted Nelson, began work on
Xanadu (196
0) focussing on hyper
-
texts in a 3
-
D space. Engelbart’s Internet
colleagues (e.g.
Baran, Kleinrock), narrowed the focus of Internet possibilities to military concerns.

A

fourth

strand applied the new model to
the
organization of
information

and categ
ories of
education. In 1964, as scientists were developing ideas of packet switching networks, the
University of Pittsburgh renamed its School of Library Science to School of Library and
Information Science.
In 1969, the year that the Internet began in the

U
.
S
,
Library Science
Abstracts w
ere renamed
Library and Information Science Abstracts.
73

The organization of
knowledge which had traditionally been the domain of learned librarians, including famous
minds such as Leibniz, was now a domain where information

science
,

as defined by computer
scientists
,

gradually acquired the upper hand.
This strand tended towards methods with
statistics and mathematical logic.
In the modern version of the Dewey Decimal System
(DDC),
000
-

Computer science, information, and gen
eral works has as a subsection: 001
-

Knowledge.
In this view, knowledge is a branch of information rather than conversely.

In the same years that information theory was being designed and written (1940
-
1948), Father
Roberto Busa (1946
-
1949), was formulat
ing a fifth strand: scholarly hypertext ideas for a new
systematic access to

knowledge in the
works of Saint Thomas Aquinas
:
hypertext for
electronic texts
avant la lettre
.
In the United States, t
his
scholarly
textual strand led to


23




Layers










Strand

Application Layer

Remote File Access

1








8

2.

Initial
Source Layer



1a.
Source (Object, Media),
Document (Text
,
Images)

1b.
Omni
-
links for letters, signs, words, images, media




(
Textual
Markup languages, SGML, XML)



5

Resource
Sharing








8

3.

Collaborative Source Layer



2a.
Studying,
Sharing






5


2b. Editing
, Sharing






2


2c. Working, Designing
, Creating

(CSCW, CSCL)



3

Directory Services








8

4.

Reference Layer


3a. Persons, Associations, Objects, Places, Dates
(Events), Processes
, Techniques, Principles


3b. Switching Layer (Top Level Headings)




6, 7




(Switching language, Matching, Search, IR languages)

3c. Terms Layer








7


(Terminology, Thesauri, Subject Headings, Classification Systems)





Remote File Access 2

5.

Content Layers








[8

4
a.

Dictionaries



4
b. Encyclopedias

4
c.
Titles in
Catalogues




6
d.

Full Texts of Sources

[and cited Sources]



6
f.
Interpretations (Secondary Literature, Reconstructions)



6
e.
External
Physical Sources
(
archaeological sites,

monuments,

heritage
)


Table 5
a
.
OSI 7 Layer Model
74

Integration
,

b. Expanded Application Layer



24


Cortazar’s Hopscotch multipath novel (1966), Brown University’s Hypertext Editing System
(HES, 1968), Alan Kay’s Dynabook (1968)
75

and App
le’s Hypercard (1987). It also led to
GML (1968), SGML (1986), XML (1996), Text Encoding Initiative (TEI, 1994) and
gradually to Digital Humanities.

Meanwhile, a sixth strand entailed philosophers exploring problems of ontology and
categories. Nikolai Har
tmann (1940, 1942, 1943), formulated a new theory of categories,
which
James K. Feibleman (1951, 1954, 1965) developed into a theory of integrative levels.
These developments were taken up by a seventh strand of classification and knowledge
organization.
T
he Classification Research Group (CRG, 19
52) was founded “to study the
theoretical foundations of classification.”
76

Meanwhile, a Broad System of Ordering (BSO)
“commissioned by UNESCO in 1971 and elaborated by the FID as a ‘root classification’ was
publis
hed in 1978.”
77

This
aimed to become a switching language. A
nother
approach was
developed in Bliss Classification, 2nd ed. (BC2, 1977) and Scheele (UFC, 1977), and further
evolved by Dahlberg (1974, and ICC, 1982, 2008) and by
Gnoli et al. (ILC, 2004).
78

In
Gnoli,
strand, knowledge, rather than information or data, becomes a top level class.

Hence, at least 7 parallel traditions evolved from the work of the 1930s and 1940s: 1) a
narrow information strand (Shannon and Weaver); 2) a computer strand focussed on

collaborative work, hypermedia and augmented IQ (Engelbart); 3) literary hypertext (Nelson);
4) information science (Pittsburgh), 5) scholarly and academic hypertext (Busa), 6)
philosophical strand (Hartmann, Feibleman); 7) classification, knowledge organ
ization strand
(CRG, Scheele, Dahlberg). Strand 1 focussed on data link layer: 2
-
7 on the application layer.

The WWW represents an eighth strand. The initial paper on Information Management
(1989),
79

which led to the WWW, mentioned none of the early 20
th

c
entury work and cited
only 1 of the 7 developments (strand 3) of the previous 50 years.
80

The good news was a
system that spread worldwide. The bad news, amidst dangers of reinventing the wheel, was
that the rich visions of hypermedia (Engelbart) and hypert
ext (Nelson) were effectively
reduced to limited (unidirectional) hyperwords, while a quest for augmenting collective
intelligence became reduced to verifying the correctness of code for truples. In the new
vision, data, rather than information or knowledg
e became key. Meanwhile, a ninth strand in
the form of DNA computing is beginning to emerge.
81



8.
Integration

of Strands


In retrospect
,

the strands and challenges
can be seen
in a fresh light. A first generation of
pioneering technologists (1930s
-
1970s)
were concerned with creating a framework and a
pipeline. For them, content was ‘merely’ an (app), and the meaning of content, information
and knowledge was ‘merely’ semantics. A second generation explored multimedia (1980s
-
).
Meanwhile, a third generation

explored the app dimension from a narrow technical viewpoint

and led to
minimal
,
unidirectional, mono
-
level links.

The framework and the pipeline became
the OSI model (table 5) with 7 layers (table 5a) and
an alternative
Internet protocol suite (IPS)
with

4 layers.
82

The first generation focussed on Physical
,
Data Link
,
network and transport
layers. The
next
generations turned
to

session, presentation and application layers.


25


Universal Classes

Before 312 A.D.



7 (+3)


(7 Liberal arts, Philosophy, Law,
Medicine)

312


1599



14

1600


1944



28

1945


1999


103


2000


2010



5 (+24)


---

1
57

(18
4
)

Table 6. Universal classes as top level headings in libraries

and classification systems.



At the top in both models, is the applicat
ion Layer, also called the End User layer (“Program
that opens what was sent or creates what was sent”
83
). It includes

Directory Services
,
Network
Management
,
Remote File Access
,
Remote Printer Access

and
Resource Sharing
. The W3C
focussed on a narrow version of Remote File Access and

effectively ignored the Resource
Sharing dimension (except for Annotea)
84

and other dimensions. To achieve the new features
outlined above (§4) requires a

further integration of earlier strands
85

and an Expanded
Application Layer.


6.1. Expanded Application Layer


The initial http protocol was about one http address leading to one remote file access. Needed
in this Remote File Access is a distinction in the Source Layer between (raw) files and files

with markup (e.g. SGML, Omnilink
). Needed in the Resource Sharing or collaborative source
layer are new collaborative tools developing the ideas of Engelbart and Nelson. Next there is
a need to revise the concept of directory services.


6.1.2. Directory Services


The OSI developed a vi
sion of a global directory service (X.500) with a Directory
Information Tree (DIT).
86

This included two subsets: s
elected attribute types
(X.520) and
s
elected object classes

(X.521). Initial use of X.520 and X.521 was for people and
associations with commer
cial applications: e.g. white pages and yellow pages. Needed is an
approach that integrates the vision of directory services with older traditions of library
catalogues. Hence the X.520 application to people could be extended and refined to include
authors
, organisations and various other names in library classifications and subject
catalogues. The X.521 category (selected object classes) could be extended to include objects
(cf. RFID), titles, places, events (dates, timelines, history), processes, techniqu
es, principles
and theories. Thus, the X.520 and X.521, which were directories of who and what
(organizations), would become directories of who, what, where, when how and why. Such a
global directory service is an excellent long term goal.


6.1.3. Switchin
g Layer

Meanwhile, the current, short
-
term reality includes many distributed, different and frequently

proprietary directories. The Broad System of Ordering (BSO, 1972) addressed this problem:



for the purpose of interconnection of information systems in the framework of the UNISIST
programme, design and develop a broad subject
-
ordering scheme, which will serve as a


26


1
.
Terms

2. Definitions

3. Explanations

4. Titles

5. Partial Contents

6. Full Contents

7. Internal Analyses

8. External Analyses

9. Conservation

10. Reconstructions

Table
7
. Levels of Knowledge


6.1.3.1.
Top Level Headings
and
Top Level Domains


switching mechanism

between

information systems and services using diverse indexing/

retrieval languages...

87

BSO, also termed SRC (Subject
-
field Reference Code) became one of
a series of systems and projects which also laid claim to being “the” switching language.
88

Some assumed tha
t the switching language could enable wholesale, simple merging between
databases. This proved overoptimistic. S
witching language
led to a series
of variants
including matching language, search language and Information Retrieval (IR) languages.

At a more
basic level, the Information Coding Classification (ICC, appendix 5a) offers and
excellent switching level for Top Level Headings (TLH) and basic concepts. For instance,
ICC 11 is Logic, which corresponds to Class Logic in TLH with equivalents such as Logi
ca in
Leibniz’ system at the Herzog August Bibliothek; a near equivalent in Bliss: Philosophy and
Logic, and a see also in Dialectic of the classical trivium.


The Top Level
Headings (TLHs)

of libraries
represent a remarkably stable field with less than
2
00 terms over the past 2000 years of which more than half have arisen in the past century
(table 6, Appendix 3).
Linking these systematically would be an excellent step in basic
interoperability between systems.



I
n the U.S. Internet, Top Level Domains (TLDs) initially
entailed
only four domains:
education, military, government and commercial (
.
edu,
.
mil,
.
gov,
.
com). They have since
been expanded to include 18 further
categories
89

as well as
c
ountry code top
-
level d
omains
(ccTLDs),
i
nternationalized country code top
-
level domains and some test tlds for major
languages.
Also planned is a GeoTLD.
90

A co
-
ordination between TLHs and TLDs would be
a
major
contribution to inte
r
connectivity.
91




6.1.3.2.
Terms Layer












These switching languages
,

especially in combination with the universal classes of Top Level
Headings,
have
t
hree

further use
s. First, they can form a bridge to
a rich array of reference
tools including t
erminology

books, t
hesauri,
s
ubject
h
eadings,
and

cl
assification
s
ystems

found in memory institutions, especially libraries, have produced
.
Second, t
hey can be linked

with top level domains of electronic resources such as BUBL
92
.

Third,
they can lead to
authority names that serve as an intermediary step i
n accessing the content layers of libraries.
For instance, a person is reading an online book which is omni
-
linked (i.e. every word is


27


bio
-
, bi
-
,
-
bia,
-
bial,
-
bian,
-
bion,
-
biont,
-
bius,
-
biosis,
-
bium,
-
biotic,
-
biotical

anima
-
, anxi
-
, deliri
-
, hallucina
-
, menti
-
, moro
-
, noo
-
, phreno
-
, psych
-
, thymo
-

anthrop
-
, anthropo
-
,
-
anthrope,
-
anthropic,
-
anthropical,
-
anthropically,
-
anthropism,
-
anthropist,
-
anthropoid,
-
anthropus,
-
anthropy

cogno
-
, meta
-
,para
-

neuro
-
, neur
-
, neuri
-
,
-
neuroma,
-
neurotic,
-
neurosis,
-
neuron,
-
neural,
-
neuria

nom
-
, nomen
-
, nomin
-
,
-
nomia,
-
nomic

nous
-
, nou
-
, noe
-
, noes
-
, noet
-
,
-
noia

Table 8. Sample
prefixes linked with the Bio area (4) and Human area (5).
93


hyperlinked
). When a word is chosen, instead of a simple link to another site, there are a
series of options in terms of content layers. The list can be an elementary (table 5b) or more
comprehensive (table 7). In either case, the implication is that univalent links
are insufficient.
Needed is a second level of remote file access.


6.1.3.3.
Remote File Access 2


This second level of remote access to files has a series of functions relating to more
information about the initial source in the remote file access 1.
Simple examples include
access to dictionaries to define a term and encyclopaedias to further explain a term. At a next
level, a reader may wish to find articles and/or books on the term. In some cases, a reader may

wish to read abstracts or reviews of a
rticles and monographs prior to deciding whether they
are relevant to the research.

Or the reader may wish to check the full text of sources cited in
the work they are reading; examine different interpretations of a text and possibly to go
beyond the writt
en sources back to the original archaeological site, monument, inscription or
other heritage site.
Sometimes they may wish to consult conservation materials concerning the
subject, or see reconstructions
.


Multilayer Links

In these scenarios the
links are

multilaye
r, potentially systematic
links to all the resources of
memory institutions pertinent to the word or claim at a series of content layers.
In future,
multilayer links could become a built
-
in feature of internet architecture. In the interim, it is
still possible to develop multilayer links without a complete reengineering of all internet links.
The current mono
-
link system can link to templates with a series of alternatives (e.g. table
7
).
These templates serve both as lists of see also

terms

to inc
rease the range of search or as lists
of filters to narrow the range of search.


If, for example, a text in Remote File Access 1 (Initial Source Layer) has the word
ethology
,
clicking
or touching the 1. Dictionary
option links the word ethology to a dicti
onary

in order
to provide a basic definition. The system would have at least three levels
:
everyday, study and
research
. The research level might begin with the same template but then offer sub
-
templates
for different dictionaries, encyclopedias etc. Initi
ally, these links can be “on the

fly
,
” simply
taking users to appropriate resources. Needed in the longer term is a harmonized,

distributed

28


resource which provides comprehensive bibliographies for individual persons, disciplines,
concept
s
, terms, words and

even letters.





6
.3.
Linkology

In the W3C
,

linking was potentially anything with everything, theoretically one to many,
practically one to one other of the many. Veracity was in the links between truples
, in the
container (pipeline) rather than the content.
In the new vision,
the role of the links
is
profoundly different.
They incorporate the perspective facet of the ILC.
The links take us
back to the sources mentioned in the text, document or source of r
emote file access 1. They
are a
n important tool for finding the meaning and context of the words and claims in our
source.
They are
also a
key to checking whether the claims made in remote file

access 1 are
identical to the sources found in remote file acc
ess 2. If there is no match, then the claims are
untrue. If the sources mentioned do not lead back to real sources that can be checked the
claims are empty. Thorough links are fundamental in a verification process. Linkology leads
to
veracity and
ontology.




7.

ICC
, ILC

and KCC (Knowledge Component Classification)

T
he
power of the
ICC (appendix
5
a)

is that its “
main structure is based on ontical levels (and
not on disciplines as all previous systems) and its divisions in the integrated levels
[are based]
o
n the so
-
called Aristotelian categories, now facets
.

94

I
n this sense it is not outdated at all

and
represents a f
undamental
snapshot of how knowledge was classified in the latter 20
th

century.
Even so, as
a model
for
the
structure
s

of knowledge
,
95

it
would benefit from
temporal
-
spatial dimensions. For instance, none of its categories provides an
exact match for
the 7 liberal arts

of antiquity.
96

In the 17
th

century the advent of the telescope and microscope
literally made visible new categories and doma
ins of knowledge.
Even in the baroque period,
many of the
ICC
categories (cf.
disciplines
)

did not yet exist
: in
16
8
2,
statistics, cybernetics,
microbiology, information science, computer science,

c
ommunication engineering and
semiotics
were not even emergent sciences. Today, a mere 30 years later, there are
new
categories and
disciplines absent from the ICC:
new neuro
-
, cogno
-

nano
-
, pico
-

disciplines,
and
trends of convergence between NBIC technologies (nano
-
, bio, info
-
, cogno
-
)
: indee
d,
a
whole range of new knowledge as scientists explore
in detail
scales from
10
-
6

(neuro
-
) to
10
-

9

(nano), 10
-
12

(pico)
and smaller
.
Needed is an expanded, temporal
-
spatial version of ICC that
illustrates the evolution of
ontical concepts,
fields of know
ledge

which are mapped to
d
isciplines
.


KCC

An alignment between
ICC
,
knowledge prefixes and the main areas of the ILC (
A
ppendix
5

b)

was noted earlier. The basic framework of the IIC
offers a framework for
understand
ing
an
underlying system that led to
naming and ordering of disciplines in Western knowledge

Appendix
5

c)
. The form concepts
(0)
serve as root

(areas)
: eg. physis (
φύσις
)
.
Theories and
Principles
(01) entail the discipline: e.g. physics. The
O
bject
C
omponent
(02) provides sub
-
disciplines: a combination of root and areas, e.g. chemical physics, astrophysics,

29


cosmophysics, geophysics, biophysics. Activity,
P
rocess (0
3
) generates verbal prefixes: e.g.
physio
-
, chemo
-
. Property Attribute (0
4
) generates a series of

further subdisciplines. For
instance the four elements (
a
ero
-
,
p
yro
-
,
h
ydro
-
,
g
eo
-
) as prefixes combine with disciplines to
produce:
a
ero
-
p
hysics,
p
yro
-
p
hysics,
h
ydro
-
p
hysics,
g
eo
-
p
hysics.
Persons or Contd (0
5
)
leads
to names of professions: e.g. physici
st, Institutions or Contd (06) leads to Institutions and
Associations: e.g. Institute of Physics.
The final three categories (07, 08, 09) pertain to
production, application and distribution.


Two of the initial 7 liberal arts dealt with heaven and earth:
astro
-
nomy (laws of the stars) and
geo
-
metry (measurement of the earth). A
n expansion of
the categories of disciplines
expanded,
led to
further branches of knowledge
:
-
nymy (names),
-
nomy

(laws),
-
logy
(science) and
-
graphy (descriptive science). A basic p
refix such as earth (geo
-
) potentially
now led to at least five disciplines: geo
-
nymy, geo
-
nomy, geo
-
logy, geo
-
graphy, geo
-
metry.
The basic prefixes also expanded dramatically (e.g.
table 8
).
Scales of knowledge led to
further sub
-
disciplines, micro
-
physics, neuro
-
physics, nano
-
physi
c
s, pico
-
physics, femto
-
physics.
In a future system
,

a matrix of prefixes and suffixes can provide a Knowledge
Component Classification (KCC), which can serve as an

orientation in categories of
knowledge and also provide a new kind of switching “language” for subjects, classes of
knowledge.



8
.
Conclusions

Initial
visions of the Internet were about complete access to all knowledge.
Part one of the
paper examin
ed a series of compromises made for pragmatic reasons

(§1
-
2). Underlying these
compromises is a focus on who and what (entities) and a tacit assumption that all statements,
claims are universally true. This assumption, common in the field of pure science,
does not
extend to the human sciences where spatio
-
temporal dimensions include
ruined, restored,
destroyed, lost and occasionally falsified sources (§3).
Needed is a fuller approach that treats
who as living entities separate from what

(bio
-

separate from
onto
-
)

and includes
determinative and ordinal relations: where, when, how, and why, which are basic aspects of
human life and knowledge.

Th
e core of the
paper
(§4)
outlines a new approach to linking knowledge in four stages: 1)
connecting letters, words an
d terms with their particulars: attributes and relations; 2
-
3) linking
these with their sources and with alternative sources, 4) linking these with questions such that
personal (who), geo
-

(where), temporal (when), conditional (how) and causal (why) subset
s
can more readily be found.
Challenges to this vision in terms of privatisation and greed were
explored (§ 5).



The latter part of the
paper (§6) returned to the spectrum of data, information and knowledge.
T
he early
digital
pioneers added two
lower levels and removed the final stage of the spectrum
to produce a new model: b
i
t
s
, bytes, data and information.
They represented but one of eight
strands that evolved in the 20
th

century. One current challenge lies in a greater integration of
these str
ands. This entails amendments to the OSI model with 7 layers
, namely,
an expanded

30


application layer with differentiation in the remote file access

and

more tools for resource
sharing
. Needed are

new directory services

which expand beyond the who and what o
f white
and yellow pages, to include the categories of library classifications and catalogues:
i,e.
directories for where, when, how, why.

Needed also
is
a
Remote File Access 2

to link
an

initial source with reference materials and cited sources.


This im
plies a new series of
multilayer links

which can be achieved via intermediary
templates (multiple decision paths) either to expand or to filter the range of a given term
.

It
also implies (§6) a new kind of multistage linking from an initial source, to bibl
iographical
sources and potentially back to the original sources that inspired them (e.g. cultural heritage
object, monument, archaeological site). The th
o
roughness of such integrated links
offer new
tools for assessing and judging the veracity of claims.


Classification systems provide us with snapshots of knowledge categories
which change over
time
.
For instance, the Dewey Decimal System has seen 23 editions since 1876.
97

This static
dimension remains even in recent systems with integrative levels (BSO, I
CC, ILC). Needed is
a dynamic
version
of disciplines and fields of knowledge over time (and place). Here

(§7),
a
refinement and expansion of the framework in the ICC
,
linked with knowledge prefixes and
suffixes, can lead to a
Knowledge Coding Classificatio
n (KCC), which provides a
history of
knowledge concepts
,
and
offers
a further switching language among classification systems
and thesauri.
Visualization
s thereof
can
give

insights
in
to patterns in emerging fields of
knowledge.


C
urrent
semantic
web
systems
link entities and attributes
, providing
containers and pipelines
for information
,

independent of the meanings of contents.

A meaningful web of knowledge
requires systematic access to the meanings of contents. Anyone can make claims which may
or may

not be true. Multilayer links give new parameters for verifying sources and further
criteria for truth, pointing to linkology as a new tool and possibly a new discipline. Links are
good.
L
inks to true sources
are better. True links
are be
st
.


Acknowledg
ements

I am grateful to Internet pioneers: Paul Otlet, Oscar LaFontaine, Doug Engelbart, Ted Nelson,
Vint Cerf,
and Sir Tim Berners Lee.
Special thanks go
to Professor Ingetraut Dahlberg,
founder of the
G
esellschaft für
K
lassifikation

and the International Society for
Knowledge
Organisation (ISKO),
whose pioneering work on
classification, e.g.
ICC
,
is a continuing
inspiration. This essay is dedicated to her. In addition, I thank friends whose encouragement
gives me strength:
Rob
Aalders (Heerlen),
Madhu Acharya (Kathmandu),
Professor Frederic
Andres (Tokyo),
Alex Bielowski (Hague),
Vasily and Alexander Churanov (Smolensk), Dr.
Jonathan Collins (Milton Keynes), Udo Jauernig (Leonberg),
Anthony Judge (Brussels),
Andrey Kotov (Smolen
sk),
Rizah Kulenovich (Karlskrona), Magister
Franz Nahrada
(Vienna),
Professor Eric McLuhan (Toronto),
Nino Nien (Maastricht),
Dr. Alan Radley
(Blackpool)
,

Carl Smith (London)
,
Professoressa Giuseppina Saccaro Battisti (Rome), Dr
Sabine Solf (Wolfenbüttel)
,
and
Dr. Marie Luis Zarnitz (Tübingen).
Finally, I am
very
grateful to Professor Francisco Ficarra for both encouragement and publication of this paper.



31


Appendix 1. Exam for Seniors.

1) How long did the Hundred Years' War last?

2) Which country makes Pan
ama hats?

3) From which animal do we get cat gut?

4) In which month do Russians celebrate the October Revolution?

5) What is a camel's hair brush made of?

6) The Canary Islands in the Pacific are named after what animal?

7) What was King George VI's first

name?

8) What color is a purple finch?

9) Where are Chinese gooseberries from?

10) What is the color of the black box in a commercial airplane?


Remember, you need only 4 correct answers to pass.

Check your answers below ....

A
NSWERS TO THE QUIZ

1) How long did the Hundred Years War last? 116 years

2) Which country makes Panama hats? Ecuador

3) From which animal do we get cat gut? Sheep and Horses

4) In which month do Russians celebrate the October Revolution?November

5) What is a camel's hair br
ush made of? Squirrel fur

6) The Canary Islands in the Pacific are named after what animal? Dogs

7) What was King George VI's first name? Albert

8 ) What color is a purple finch? Crimson

9) Where are Chinese gooseberries from? New Zealand

10) What is the color of the black box in a commercial airplane? Orange(of course)

What do you mean, you failed?






32


Appendix 2
.

Grammar vs.
T
r
uples

(
Tuples,
Tr
i
ples) and the Semantic Web


In grammar and logic
subject, verb, object and p
redicate ha
ve

a v
ery specific meaning:


Predicate

1
.
Grammar

One of the two main constituents of a sentence or clause, modifying the subject and
including the verb, objects, or phrases governed by the verb, as
opened the door

in
Jane opened the
door

or
is very sleepy

in
The child is very sleepy.


2.
Logic

That part of a proposition that is affirmed or denied about the subject. For example, in the
proposition
We are mortal, mortal

is the predicate.
98


In the Resource Description Format (RDF)
, the meaning of subject, verb and object are
changed. Here, the verb becomes the predicate: e.g.


We put these individual pieces together to form RDF statements, which are like English sentences. RDF
statements are also pretty simple: they have a
subject (the thing you're talking about), a predicate (what
you're saying about it), and an object (the thing you're saying). For example, take this English sentence:

My widget has the title "Mega Widget 2002".

"My widget" is the subject, "has the title" i
s the predicate, and "Mega Widget 2002" is the object. Here's
that same RDF statement in N
-
Triples:
99

This alternative form of grammar
is further discussed in an introduction to the Sematic Web
for laymen:

The Semantic Web is a set of standard technologies for modeling information. They can be applied to
almost any problem.

The Data Model of the Semantic Web is RDF (Resource Description Framework). An item in RDF is 3
-
tuple (Subject, Predicate, Object), and
3
-
truples connect to form a Graph. There are RDF Databases
(aka "Triple Stores"). You can think of this as a form of NoSQL Database; extremely flexible in its
ability to store information as compared to a relational database or XML.

Data in RDF is describe
d via OWL (acronym for Web Ontology Language...yes the O and W are mis
-
ordered) ontologies. An "Ontology" is a fancy word for "Data Model." You use ontologies to describe
data. In this way, Semantic Web data modeling is similar to duck typing; data exists,

and ontologies
describe the data that exists. One man's Terrorist may be another man's Freedom Fighter, for example.
For two applications to exchange information, they have to agree on ontologies (though merging data
from two ontologies is very much easie
r than the ETL work required to merge data from multiple
databases).

The Query Language of the Semantic Web is SPARQL. It is designed to query distributed graphs of
information (e.g. if data is distributed across multiple RDF stores, you can query across t
hem
seamlessly from a single SPARQL query, which is a HUGE difference as compared to SQL or XQuery,
for example).

The hype around this stuff is this "world wide database" or "Linked Data Cloud" vision, whereby all
information in all places is tagged semantically so can be queried across, merged, and analyzed at will.
Some progress has been made towards this end (GoodD
ata, Schema.org, etc.), but the promise still
seems distant.
100





33


Appen
d
ix 3.
Universal Classes (Top Level Headings

Before 312 A.D
101
.





7

Class: Arithmetic

Class: Astronomy

Class: Dialectic

Class: Geometry

Class: Grammar

Class: Music

Class: Rhetoric


312
-

1599 A.D.






14

Class:
Arts

Class: Biography

Class: Economics

Class: Ethics

Class: General

Class: Geography

Class: History

Class: Logic

Class: Manuscripts

Class: Physics

Class: Poetry

Class: Politics

Class: Theology

Class: War


1600
-
194
4

A.D.






28

Class: Agriculture

Class: Arts

Class: Auxiliary Sciences of History

Class: Bibliography

Class: Books

Class: Education

Class: Fine Arts

Class: History: America

Class: Jurisprudence

Class: Language

Class: Language and Literature

Class: Law

Class: Library Classifications

Class: Library Science

Class: Linguistics

Class: Literature

Class: Mathematics

Class: Medicine

Class: Military Science

Class: Natural Sciences

Class: Other Applied Sciences

Class: Philosophy

Class: Psychology

Class: Religion

Class: Scholarship

Class: Science

Class: Social Sciences

Class: Technology

Class: Technology (Applied Sciences)




34


1945
-
1999

A.D.





1
03

Class: Art Sciences

Class: Biology

Class: Book Science

Class: Books on M
usic

Class: Botany

Class: Business Administration

Class: Business Administration, Organizational Science

Class: Chemical Engineering

Class: Chemistry

Class: Civil Engineering

Class: Civilizat
ion

Class: Classical Mythology

Class: Communication Studies

Class: Computer Science

Class: Concepts

Class: Criminology

Class: Cultural Anthropology

Class: Culture

Class: Demographics

Class: Documentary Information

Class: Domestic Science

Class: Dramaturgy

Class: Earth Sciences

Class: Electrotechnology

Class: Engineering

Class: Environmental Science

Class:

Ethnology (of non
-
European cultures)

Class: European Ethnology

Class: Exact Sciences in General

Class: Fine Arts

Class: Folklore

Class: Forestry

Class: Gender Studies

Class: Genetics

Class: Geology

Class: Health Sciences

Class: History Europe Asia Africa

Class: Human

Class: Human Being, Man

Class: Human Biology

Class: Human Environment

Class: Humanities in General

Class: Information

Class: Information Resources

Class: Information Science and Technology

Class: Information Sciences

Class: Journalism

Class: Knowledge

Class: Languages

Class: Leisure Activities

Class: Linguistics of Separate Languages

Class: Literary Studies

Class: Literatures

Class: Magic

Class: Management of Economic Enterprises

Class: Maps

Class: Materials Science

Class: Mechanical Engineering

Class: Mining Engineering


35


Class: M
orals

Class: Music Science

Class: Musicology

Class: Nature

Class: Naval Science

Class: Occult

Class: Organizational Science

Class: Pedagogy

Class: Phenomena

Class: Physical Anthropology

Class: Physi
cal Education

Class: Political Science

Class: Political Sciences

Class: Probability

Class: Process Technology

Class: Psychiatry

Class: Publi
c Administration

Class: Recreation

Class: Religious Studies

Class: Research

Class: Research and Scholarhsip

Class: Science and Culture

Class: Science of Public Administra
tion

Class: Separate Art Forms

Class: Social Anthropology

Class: Social Geography

Class: Social Science

Class: Social Sciences in General

Class: Social Welfare

Class: Society

Class: Sociology

Class: Space Sciences

Class: Statistics

Class: Structure

Class: Teaching

Class: Technical Science

Class: Theory of Adult Education

Class: Thought

Class: Traffic

Class: Transport Techn
ology

Class: Travel

Class: Veterinary Medicine

Class: Veterinary Science

Class: Virology

Class: Zoo
logy


2000 A.D.






5

(+ 24)

Class: Books on Music

Class: Chemical
Engineering

Class: Information Resources

Class: Naval Science

Class: Political Science

Top Level Headings

Universal Classes

class: a. form

class: b. spacetime

class: c. energy

class: d. particle
s

class: e. atoms

class: f. molecules

class: g. bodies


36


class: h. celestial objects

class: i. weather

class: j. land

class: k. genes

class: l. bacteria

class: m. org
anisms

class: n. populations

class: o. instincts

class: p. consciousness

class: q. signs

class: r. languages

class: s. civil society

class: t. governments

class: u. economies

class: v.
technologies

class: w. artifacts

class: x. art

class:
y. knowledge

class
:
z.religion




Appendix
4
.
Integrative Levels Classification

(ILC)
:
main classes
:
(select
to expand)

aligned
with basic questions

Who


What


Where




When


How


Why

a.

form

b. space



b. time

c.energy

d. particles



e.atoms



f.molecules

g.bodies





h.celestial objects

i.weather



j.land



k.genes



m.organisms

n.populations

n.populations

o. instincts

p. consciousness

q.signs

r.languages

s.civil society

t.governments

u.economies

v.technologies

w.artifacts

x.art

y. knowledge

z.religion





37


Appendix
5
.
a.
Ingetraut Dahlberg, Information Coding Classification (ICC)
, b.
ILC
, c. KCC
.



What a

What
[

b
]


What

c


What d

What
e


Who a

Who b

How a

How b

How

c



= Why


Nine Areas



Prefixes



ILC

1.Form & Structure



phylo
-
, morpho
-


a.form

2.Energy & Matter



E, hylo



c.energy,
d.particles,e.atoms,f.molecules, g. bodies

3.Cosmo & Geo
-




cosmo
-
, geo
-


b.spacetime,

h.celestial objects, i.weather,
j.land

4.Bio
-




bio
-




k.genes,

m.organisms, n.populations

5.Human




anthro
-



o. instincts, p. consciousness

6.Socio




socio
-





s.civil society, t.governments

7.Economics & Technology


econo
-
, techno
-


u.economies,
v.technologies

8.Science & Infor
mation

scientific, info
-
, cogno
-



y. knowledge

9.Culture


cultur
o
-
, cultural


q.signs, r.languages, w.artifacts, x.art,
z.religion




What a

What[ b]


What c


What d

What e

Who a

Who b

How a

How b

How c




= Why

Root




Discipline

Subdiscipline
Activity

Property

Person


Institution


P
roduction
A
pplication

D
istribution






-
er,
-
or,
-
ist

physis physics


---
physics
physio
-


---

physics physicist Institute of physics


chem


chemistry

chemi
cal




chem
o
-


chemical
--


chemist

Institute of Chemistry



chemology


nomos nomothethic nomo
-

nomethetical
νομοθέτης

(lawgiver )

logos

-
ology


logic


logico
-

logical logician

graph


-
graphy


graphic


grapho
-

-
graphical


-
grapher



Institiute of
-
graphy


-
graphical

-
graphical
-
graphical












production application distribution

metr
-



geo
-

logos


geology

geology

geological geologist
Institiute of geology

geo
-

graph


geograph
y


geograph


geographic

geographical

geographer

Institute of


geography







38





Notes

1

See also: Timeline of Systematic Data and the Development of Computable Knowledge:
http://www.wolframalpha.com/docs/timeline/computable
-
knowledge
-
history
-
5.html

2

Bits and Bytes: http://computer.howstuffworks.com/bytes.htm

3

Shannon: http://en.wikipedia.o
rg/wiki/Claude_Shannon

he is also credited with founding both
digital computer

and
digital

circuit

design theory in 1937, when,
as a 21
-
year
-
old
master's degree

student at the
Massachusetts Institute of Technology

(MIT), he wrote
his
thesis

demonstrating that electrical applications of
boolean algebra

could construct and resolve any
logical, numerical relationship.

4

Cited from: Claude Shannon, Warren Weaver, "A Mathematical Theory of Communication":

http://www.uoregon.edu/~felsing/virtual_asia/info.html

5

Just a matter of semantics:
http://sandradodd.com/semantics
;
http://english.stackexchange.com/questions/97318/is
-
the
-
phrase
-
its
-
just
-
a
-
matter
-
of
-
semantics
-
meaningless

6

The basic distinctions between subsumptive, deteminative

and ordinal relations were developed by Jean
Perrault (Boca Raton, Florida International University), 1965, in conjunction with Ingetraut Dahlberg. See : Jean
Perrault, “Categories and Relators,” International Classification, Frankfurt, vol. 21, no. 4. 19
94, pp. 189
-
198,
especially p. 195. Cf.
Jean M Perreault
, Towards a theory for UDC; essays aimed at structural understanding and
oper
ational improvement, [Hamden, Conn.] Archon Books [1969].

7

Apple created a software, Hypercard (1987), which made basic aspects of the process accessible to everyday

users but then abandoned the product.

8

XML Timeline: http://www.dblab.ntua.gr/~bikakis/X
MLSemanticWebW3CTimeline.svg

9

TCP: http://www.textcreationpartnership.org/tcp
-
eebo/

10

ECCO
-
TCP:
http://www.textcreationpartnership.org/tcp
-
ecco/

:

The database contains more than 32 million
pages of text and over 205,000 individual volumes in all. In
addition, ECCO natively supports OCR
-
based full
-
text searching of this corpus.

ECCO
-
TCP

With the support of more than 35 libraries, the TCP keyed and encoded 2,231 ECCO
-
TCP texts. In
cooperation with Gale Cengage, these texts have already been made freely available to the public

11

Tim Berners Lee, Information Management: http://www.w3.org/His
tory/1989/proposal.html

12

Ibid: “I imagine that two people for 6 to 12 months would be sufficient for this phase of the project.”

13
Ibid:
People, Software modules, Groups of people, Projects, Concepts, Documents, Types of hardware,
Specific hardware object
s.

14

These offered an alternative to the Transfer Control Protocol/Internet Protocol (TCP/IP) which underlay the
Internet.

15

Some sites included a redirect function whereby searching for earlier site
a
led automatically to new site
b
.
Alas almost no sit
es offer information or forwarding for former sites which are now defunct. An exception are
some sites formerly on Geocities.com (now defunct), which are now maintained at ReoCities.com. Internet
Archive also has records of many no longer extant sites as
does Google.

16

HTML Timeline: http://topshelfcopy.com/wp
-
content/uploads/2012/12/html
-
timeline.png

17

Usability glossary: http://www.usabilityfirst.com/glossary/hypertext/

18

Semantic Web:
http://www.reddit.com/r/semanticweb/comments/ksykt/im_new_and_what_is_this/

19

For the humane sciences the above statements are too imprecise. Was it John the Baptist,

Pope John XXIII,
John Lennon or John the neighbour’s boy whom most call Johnny? When and where did John walk? How did he
walk? Why did he walk?

20

Ontology: http://wenku.baidu.com/view/6c574f1dc5da50e2524d7f01.html

? A systematic account of existence.

? What it means to exist ? Deals with order and structure of reality
Ontology


Definition (AI, CS) ? Multiple definitions have been coined ? That which exist ? that which can be
represented ? An explicit specification


>>

21

Tim Berners Lee, Information Management (1989/1990):
http://www.w3.org/History/1989/proposal.html
:

and

yes, this would provide an excellent project with which to try our new object oriented programming
techniques!

22

Internet of Things: http://upload.wikimedia.org/wikipedia/commons/5/5a/Internet_of_Things.png

23

These results were on 13 February 2013.

24

Visu
al Particulars.
The visual world is different. A word is universal. A photographic picture is individual. The
word dog applies to all dogs in all times and all places. A photograph provides an image of 1 dog in 1 place at 1
time. In terms of that specific
dog (e.g. in London at 9 a.m. on 1 February 2013) a picture may be worth 1000
words. It may convey a sense of dogginess, but can give us little idea of the range of sizes from miniature

39







chihuahuas to great Danes, dogs in China, or dogs in the Roman Empire.

And unless the details of the precise
place and time are recorded, it will in future often be impossible to identify the exact time and place, except if
there clues in the picture itself such as Big Ben with its clock striking 9.

These characterist
ics change with different media. For instance, movies sometimes record physical places at a
specific time. They also combine “real” elements in ways that are no longer a match with the physical world. For
instance, Schloss Adler and the funicular in the fi
lm Where Eagles dare, reflect two distinct places which are not
linked in the physical world, namely: Burg Hohenwerfen at Werfen and
Feuerkogel Seilbahn at
Ebensee
, both in
Austria. Cf. Wher
e Eagles Dare: http://en.wikipedia.org/wiki/Where_Eagles_Dare

25

On 21 02 2013

26

See: Derrick De Kerckhove: http://www.40kbooks.com/?p=3811

27

Data Link Layer:
http://en.wikipedia.org/wiki/Data_lin
k_layer

.This is level 2 in the OSI model.

28

Link Layer: http://en.wikipedia.org/wiki/Link_layer

29

It is striking that the English search for elementary particle in GBV produces an entirely different set of
keywords:
tsotsas
,
swarm
,
kharaghani
,
sunkara
,
fluidized
,
granulation
,
polyacrylamid
,
nanotechnology
,
astrophysics
,
partikeltechnologie

30

EZB: http://rzblx1.uni
-
regen
sburg.de/ezeit/searchres.phtml?bibid=SUBGO&colors=7&lang=de&jq_type1=KT&jq_term1=elementary
%20particle

31

EZB, Göttingen: http://rzblx1.uni
-
regensburg.de/ezeit/searchres.phtml?bibid=SUBGO&colors=7&lang=de&jq_type1=KT&jq_term1=elementartei
lchen

32

Scitation:

http://scitation.aip.org/vsearch/servlet
/VerityServlet?KEY=FREESR&possible1=elementary+particle&possible1
zone=article&bool1=and&possible2=&possible2zone=multi&bool4=and&possible4=&possible4zone=author&p
ossible_adv=&sort=chron&maxdisp=25&threshold=0&frommonth=&fromday=&fromyear=&tomonth=&today
=&t
oyear=&fromvolume=&fromissue=&tovolume=&toissue=&smode=strresults&ver=&sti=&page=1&origque
ry=&vdk_query=&chapter=0&docdisp=0&%5Bsearch%5D.x=0&%5Bsearch%5D.y=0

33

For another discussion of these problems with more attention to the layers see the author’s 200
5 Access,
Claims and Quality on the Internet


Future challenges, Progress in Informatics, Tokyo, no. 2, November 2005,
pp. 17
-
40: ttp://sumscorp.com/new_media/computers/internet/news_161.html

34

Aristotle, Categoriae (chapter 4): http://classics.mit.edu/Ar
istotle/categories.html

35

In Dahlberg’s approach, substance is treated as having 9 accidents. Dahlberg
enlarged “the single ‘substance’
into three kinds and with the nine accidents…found that there are three properties, three activities (although one
activity is static) and three dimensions” and thus “created the 4 ur
-
categories of which the Aristotelian ones are
then subdivisions and thus facets.” (personal communication).

36

John: http://www.helium.com/items/770850
-
behind
-
the
-
name
-
john

37

Alans:
http://www.facebook.com/note.php?note_id=10150318997685145

Subdivisions and ethnic affiliates
Alans, Burtas, Rhoxolani, Wusüns, Yasses, Yazygs

38

Scythia et Serica:
http://en.wikipedia.org/wiki/File:Scythia_serica.jpg

Sarmatia et Scythia:
http://www.bergbook.com/images/24010
-
01.jpg

Russia:
http://www.lib.utexas.edu/maps/commonwealth/commonwealth.jpg

39

Tim Berners Lee wrote about the problem with keywords:
http://www.w3.org/History/1989/proposal.html

Ironically,

40

Aristotle’s defin
ition of reality is very different than the contemporary one. For him relation is a comparative
term: whether an object is greater than, smaller than etc.

41

Aristotle, Categories : http://plato.stanford.edu/entries/aristotle
-
categories/

42

Determinative
Relations
:


Acting


Being acted upon

Active


Passive




Operations

Processes


Efficient Cause

Material Cause

43

E.g. Calendar conversion: http://www.fourmilab.ch/documents/calendar/

44

Vasistha and Zoroaster:
http://www.topix.com/forum/religion/zoroastrian/THKNUB3S16PAT480T

:


40







According to the Vedic version, Zoroaster and Vasistha were half brothers. Vasistha was the legitimate
son o
f Surya and Zoroaster was the illegitimate son of Surya and the maiden Niksubha. In their adult
lives both Vasistha and Zoroaster became priests of Asura Varuna [possibly in Kashmir]. Vasistha and
Zoroaster were co
-
priests of Varuna but in due course there

would arise irreconcilable differences
between the two. So great was the rivalry between Vasistha and Zoroaster that the latter eventually
separated himself from the Vedic standards. Zoroaster gathered his followers and made an exodus
toward the west, eve
ntually settling in Persia [north
-
eastern Iran]. This new religion of Zoroaster was
more like a rehashing or mixing of the old Vedic beliefs with an occasional addition of his own.
Zoroaster took the concepts of gods and demons found in the Vedic pantheon
and reassigned them
different names and different functions. From among those Zoroaster favored Varuna whom he called
'Ahura Mazda', the Supreme God. Surya or Mitra, the Vedic sun
-
god, also took his place in the belief of
the Zoroastrians as did the worshi
p of fire. To the Persians Mitra became Mithras. Vasistha and his
followers were called Brahman and they worship the Devas Chief God Indra the Moon God or Chandra
and Drank Soma. Zoroasters and his Magas Magi and the worship Chief Asuras God Varuna the Sun

God or Surya and worship fire.

45

Zoroaster, Wiki: http://en.wikipedia.org/wiki/Zoroaster

46

These problems apply equally to thorny questions of authorship, especially in the realm of painting where a
master has students, assistants, sometimes a school and followers. Art history and especially connoisseurship has
developed a range of vocabulary t
o describe this range. Hence a painting is by x, Attributed to, Ascribed to,
Student of, Workshop or School of, Follower of, or merely a copy. Paintings in galleries typically have one of
these alternatives. Learned articles typically provide the whole his
tory of attributions. If organized in database
fashion then one could view these claims chronologically, and see how many claims are in each of these
categories with respect to a given painting.

47

ILC:
http://www.iskoi.org/ilc/1/no.php?no=8&sp=3

On the surface, the 26 categories of ILC are what cat
e
gories, with only a few as obvious candidates for other
questions: e.g. b. spacetime, i weather (cf. appendix 4). At a greater level of detail each

of the 26 categories
entails the six questions: e.g. there are names connected with form, as well events, places , theories etc.

Cf. verbs under Instincts and prepositions under Aspects.

48

ILC: http://www.iskoi.org/ilc/1/no.php?no=9&sp=3

49

Trivium (Gram
mar, Dialectic, Rherotic) and quadrivium(Arithmetic, Geometry, Astronomy, Music)

50

Unless it is a review article in major papers such as the New York Review of Books or the Sunday section of
the Frankfurter Allgemeine.

51

For an earlier treatment see the au
thor’s Reality, Knowledge and Excellence:
http://sumscorp.com/new_media/knowledge/knowledge_organisation/news_205.html

52

There were a number of versions. Some defined an empyrean beyond the fixed stars. Some linked the 7
heavens with the 7 planets and defined the intermediate space as the atmosphere as the area between the moon
(nearest planet) and the earth.

53

Chadwyck
-

Healey: http://www.proquest.com/en
-
US/products/brands/pl_ch.shtml

54

Proquest: http://www.proquest.com/en
-
US/aboutus/default.shtml

55

CIG: http://www.cig.com/

56

Bowker: http://www.proquest.com/en
-
US/products/brands/pl_bowker.shtml

57

EEBO: http://eebo.chadwyc
k.com/home

58

WBIS: http://db.saur.de/WBIS/login.jsf

59

PDA: http://www.degruyter.com/page/428

60

PDA: http://www.degruyter.com/page/428

61

Mosul looting: http://archive.archaeology.org/iraq/mosul.html

62

Terrorism and Destruction of heritage:
http://www.middleeastmonitor.com/resources/commentary
-
and
-
analysis/5026
-
frances
-
record
-
in
-
the
-
middle
-
east
-
rules
-
out
-
any
-
constructive
-
role
-
in
-
mali
;

63

In Malaysia, there is a careful distinction between knowledge and enduring knowledge which leads to a corpus
of the memorable.

64

Traité: http://archives.mundaneum.org/en/history

65

Paul Otlet, Monde: essaie d'universal
isme
--

connaissance du monde; sentiment du monde; action organisée et
plan du monde, Brussels, Editions du Mundaneum, 1935
http://www.laetusinpraesens.org/docs/otlethyp.php
;

Man would no
longer need documentation if he were assimilated into an omniscient being
-

as with God
himself. But to a less ultimate degree, a technology will be created acting at a distance and combining
radio, X
-
rays, cinema and microscopic photography. Everything in

the universe, and everything of man,
would be registered at a distance as it was produced. In this way a moving image of the world will be
established, a true mirror of his memory. From a distance, everyone will be able to read text, enlarged

41







and limited
to the desired subject, projected on an individual screen. In this way, everyone from his
armchair will be able to contemplate creation, as a whole or in certain of its parts.

66

Computing History Timeline: http://www.columbia.edu/cu/computinghistory/#timel
ine

67

Engelbart: http://www.dougengelbart.org/about/augment.html

68

Collective IQ:
http://www.dougengelbart.org/about/vision
-
highlights.html

Cf. Engelbart: http://www.dougengelbart.or
g/about/augment.html

69

Engelbart Innovations: http://www.dougengelbart.org/history/engelbart.html

70

DKRs: http://www.dougengelbart.org/about/dkrs.html

71

Open Hyper Tools: http://www.dougengelbart.org/about/open
-
hyper
-
tools.html

72

Engelbart, OHS: http://www
.dougengelbart.org/about/ohs.html

73

LIS:
http://en.wikipedia.org/wiki/Library_and_information_science

74

OSI: http://www.escotal.com/Images/Network%20parts/osi.gif

75

Hypertext: ht
tp://people.lis.illinois.edu/~chip/projects/timeline/1453hudson.html

76

CRG:
World Encyclopedia
of

Library an
d Information Services, 3rd Ed, 1993, p. 211.

77

Ibid.

Professor Dahlberg adds (personal communication):

Eric Coates was in charge, but before that he was a member of a group to work towards an SRC,

in the
FID
-
Classification Research Group and in 1974,
at a meeting in The Hague the work of this group was
given to a 3
-
man group to use the material so far elaborated (among which thousands of terms denoting
fields of knowledge, which I had brought in together with my way of arranging them what was later on
called the ICC) in order to elaborate a final version of the BSO.

78

ILC: http://www.iskoi.org/ilc/index.php

79

Op. cit,:
http://www.w3.org/History/1989/proposal.html
:

80

Not even XML (1996), which was to become a basic component of the W3C, was acknowledged.

81

An ounce of DNA will potentially allow us to store the equivalent of 1 trillion CD
-
ROMS. Technologically, it
will be possible to carry the contents of the world’s m
emory on a DNA “stick” that weighs less than the
computer sticks of today. The rhetorical need for cloud computing may prove superfluous, and a future Internet
may be focussed on updates, sharing and communication.

82

The DOD (Department of Defense
) 4 model also has 4 layers.

83

OSI: http://www.escotal.com/Images/Network%20parts/osi.gif

84

Annotea: http://www.w3.org/2001/Annotea/

85

Technical details of how this affects other aspects of the Applications Layers such as Network Management
are not our co
ncern here.

86

OSI: X.500:
http://en.wikipedia.org/wiki/X.500

This was connected with a series of ten standards (X.500
-

X.530).

87

BSO:
http://www.ucl.ac.uk/
fatks/bso/
. Cf. note 77 above.

87

The BSO has 10 basic subject headings, with categories in the range 100
-
972
87

and 6800 subjects in all.

See: BSO: http://www.ucl.ac.uk/fatks/bso/outline.htm

88
Other systems include the Vocabulary Switching System (VCC), the
International Patent Classification
catchwords ; the HILT, and RENARDUS projects.
Cf.

WIPO:
http://www.wipo.int/classifications/ipc/en/est/

See also: http://www.iva.dk/bh/lifeboat_ko/CONCEPTS
/switching_language.htm

89

TLDs: http://en.wikipedia.org/wiki/List_of_Internet_TLDs

90

GeoTLD: http://en.wikipedia.org/wiki/GeoTLD

91

For details see:
Internet Domain Names and Indexing

(2002):
http://sumscorp.com/new_media/computers/internet/news_154.html

Cf
.
Domain Names and Classification Systems

(2002):

http://sumscorp.com/new_media/computers/internet/news_151.html

92

BUBL:
BU
lletin
B
oard for
L
ibraries:
http://bubl.ac.uk/link/subjectbrowse.cfm

. Note how a majority of these
subject headings are those of the Top Level Headings of libraries.

93

See English Word Information: http://wordinfo.info/units

94

Dahlberg (Personal communication). It also has “
many subdivisions into sub
-

subsub
-
, subsubsub,
etc
-
fields in
which all those old and new fields have been and can also be incorporated.


95

Ontology in the old sense.

96

Even so, clear parallels exist:

Grammar


91. Language and Linguistics

Dialectic


11. Logic

Rhetoric



85. Communication Science

Arithmetic


12. Mathematics



42







Geometry


12. Mathematics

Music



93. Music

Astronomy


31. Astronomy


97

DDC: http://en.wikipedia.org/wiki/Dewey_Decimal_Classification

98

Free Dictionary: Predicate: http://www.thefreedictionary.com/predicate

99

RDF Primer:
http://notabug.com/2002/rdfprimer/

100

http://www.reddit.com/r/semanticweb/comments/ksykt/im_new_and_what_is_this/

In

regular grammar there is distinction between intransitive and transitive verbs. Intransitive verbs (is a) entail
entitities
-
attributes and are copulas (links, ties), have subsumptive relations and no objects. Transitive verbs (has
a, builds a) entail dete
rminative relations, (activities), which have objects: e.g. John hit the nail (with his
hammer). In RDF no distinction is made between transitive and intransitive verbs. So “John has an automobile”
and “John is a man” are treated equally as truples. No dis
tinction is made between living entities and entities. So
“Cain killed Adam” (an intentional act and a crime) is treated on equal terms qua truples as: “The car hit a tree”
(an accident, not intentional and not a crime). This “higher level of abstraction”
has the advantage that it
simplifies the model. The bad news is that it limits discussions to subsumptive relationships, with no place for
determinative or ordinal relationships within the model.

This may seem trivial but in a world where
tagging and l
inking are all the rage in a quest for an Internet of
things, it is essential to have criteria to distinguish between mere opinions (empty claims) and serious evidence;
between links to sources and links to reports about or opinions concerning sources. In
RDF all links are treated
equally. In reality, not all links are equal: some are true, some are false.


In a binary situation, true and false are a matter of yes or no, on or off, white or black. In real life, events occur in
time and space. So “Cain ki
lled Abel” becomes a claim: “Cain killed Abel at 3 pm in the cornfield.” If there be a
witness who was in that cornfield at that time, then the statement can be verified. Various conditions can also be
added: e.g. using his shovel just as the sun was going

under a cloud. Motives can also be added: because he was
jealous of being less acceptable in the eyes of God. Every further detail provides further criteria for determining
the veracity of any claim or story. This is the context of the cross
-
examinations
of Inspector Colombo, Hercules
Poirot, Perry Mason and their ilk. RDF addresses the “what” of a sentence, which assumes scientific entities. The
truth of a claim entails the who, what, where, when, how and why of a statement. For RDF, truples are sufficien
t.
For truth, sextuplets (sextruplets) are a minimum. Minimal data is in bits (single digits or letters). Minimal
information is in bytes (8 bits or words). Minimal knowledge requires much more.

Data is about bits and bytes and no questions. Information c
an extend to 2 questions: Who and What?
Knowledge is potentially about 6 questions: Who, What, Where, When, How, Why?

101

Although they were not technically a part of the 7 liberal arts, implicitly there must have been classes for
Class: Law, Class: Medic
ine, Class: Philosophy.