THE WELCH COMPANY
440 Davis Court #1602
San Francisco, CA 94111-2496
415 781 5700
rodwelch@pacbell.net


S U M M A R Y


DIARY: September 14, 2001 10:59 PM Friday; Rod Welch

Pictures and text included in organizational memory, history.

1...Summary/Objective
2...Graphical Representation Not Effective for Intelligence, KM
3...Graph Technologies Not Effective Collaboration Knowledge Management
4...Representation Selected Elements Makes Complexity Comprehensible
5...Visualization Pictures Graphics Useful Often Essential
6...Pictures Worth 1,000 Words, Instantly Conveys Information
......No Perceived Threat or Gain, Laziness Encourages Poor Work
......People Care About Fast and Easy, Not Productive Useful
.....Intelligence Demanded When People Get Hurt, Income Threatened
.....Care About Safety and Income Increase When Threats Perceived
.....Threats to Safety and Income Change What People Care About
.....Intelligence Produces Knowledge to Be Prepared
.....Organizational Memory Captured by Context of What People See Hear


..............
Click here to comment!

CONTACTS 
0201 - Deneen Consulting
020101 - Mr. John Deneen
0202 - Armstrong Consulting
020201 - Mr. Eric Armstrong

SUBJECTS
Pictures Worth 1000 Words
Pictures Worth Thousand Words, John Deneen, 010914
Information Structured Text Notes, Eugene 000504
Graphical Display Only Useful for Small Structures Overwhelmed by Com
Picture Thousand Words Explain Analyse Understand Follow Up Need Grow
Graphic Explanation Why Impractical for KM Overwhelmed by Complexity
Text Communication Meaning Inference, Semiotics, 000713
Pictures Require Thousand Words Explain Analyse Understand Follow Up

3310 -
3310 -    ..
3311 - Summary/Objective
3312 -
331201 - Follow up ref SDS 19 0001 ref SDS 10 6205.
331202 -
331203 - Eric Armstrong provides detailed reasoning that confirms earlier
331204 - analysis showing pictures and graphs are not effective for large scale
331205 - knowledge management.  John Deneen objected on the grounds that a
331206 - picture is worth a thousand words.  John submits support that says
331207 - visualization are effective for certain types of information, and
331208 - cites pictures of human faces. ref SDS 0 JF7H  The record on this is
331209 - issue is presented showing that pictures are very attractive to human
331210 - biology that is wired to use sight and sound. ref SDS 0 SL3J Eric has
331211 - previously cited the challenge that pictures require a lot of
331212 - explanation.
331214 -  ..
331215 - Submitted ref DIT 1 0001 explaining intelligence integrates pictures,
331216 - text and time.  Pictures are a vital part of human understanding that
331217 - require a lot of explanation to integrate them into the context of
331218 - daily activity. ref SDS 0 JF4O
331219 -
331220 -
331221 -
331223 -  ..
3313 -
3314 -
3315 - Progress
331601 -  ..
331602 - Graphical Representation Not Effective for Intelligence, KM
331603 -
331604 - Received ref DRT 1 0001 from Eric Armstrong commenting on a letter
331605 - from Alex Shapiro recommending review of a pdf paper.....
331606 -
331607 -
331608 -       http://www.cs.vu.nl/~frankh/postscript/VSW01.pdf
331609 -
331610 -
331611 - ...on using graphics for knowledge management.
331613 -  ..
331614 - Eric says in part....
331615 -
331616 -     The examples in this paper appear to me to reinforce the
331617 -     principles I posited in a post quite a while back. ref DRT 1 Q66I
331618 -
331619 -     Graphics work when there is...
331621 -           ..
331622 -       •  a small set of
331623 -       •  fixed data types
331624 -       •  small sets of relationships
331626 -      ..
331627 -     [On 030317 Jeff Conklin suggests dialog maps in formal meetings to
331628 -     apply IBIS, ref SDS 24 5W4K; indicates application for other
331629 -     communication is harder, ref SDS 24 5W59, because complexity
331630 -     exceeds span of attention. ref SDS 24 XU7L
331632 -  ..
331633 - Eric's analysis today aligns with his letters on....
331634 -
331635 -     Text nodes atomic data structure.......000423, ref SDS 9 4977
331636 -     Graphical Magic Lens Not Practical.....001008, ref SDS 12 N38W
331637 -     Ken Holman's presentation at SRI.......001113, ref SDS 13 3Q5M
331638 -     Picture takes 1,000 words to explain...010312, ref SDS 16 Q56M
331639 -     Graph representation not effective.....010912, ref SDS 19 0001
331640 -     Topic maps take a lot of time..........010916, ref SDS 21 4J4J
331642 -  ..
331643 - On 000504 Eugene Kim related experience showing text and free form
331644 - narrative and commentary using alphabet technology has advantages for
331645 - knowledge management, commonly called literacy. ref SDS 10 6205
331647 -  ..
331648 - Biological drive for sight and sound to obtain information requires
331649 - "intelligence" that converts information into useful knowledge,
331650 - reviewed below. ref SDS 0 SL3J
331651 -
331653 -  ..
331654 - Eric continues...
331655 -
331656 -     Graphical representation allows one icon to be associated with
331657 -     each type. The graph can then show patterns or locations of the
331658 -     items. ref DRT 1 J66O
331659 -
331660 -     Graphs run into problems in one of three ways:
331661 -
331662 -       1.  When the number of types grows large, there are too many
331663 -           icons to keep track of, and no meaningful patterns emerge.
331664 -           ref DRT 1 NF7I
331666 -            ..
331667 -       2.  When the number of relationships grows large, the
331668 -           intersecting lines in any graphic representation turns the
331669 -           picture into a confusion. ref DRT 1 8G7O
331671 -            ..
331672 -       3.  When the number of entries grows large, items are far
331673 -           removed from each other, and the other end of any given
331674 -           relationship is rarely visible in a given display area.
331675 -           ref DRT 1 AG8H
331676 -
331677 -     ....examples used in the PDF paper, ref SDS 0 UU8N, have exactly
331678 -     two data types: a location at the top level of the hierarchy, and
331679 -     something else (presumably a "job" type) at the second level of
331680 -     the hierarchy. I note that no information about the job is
331681 -     contained in the graph. So the "information content" only goes one
331682 -     level deep. ref DRT 1 UG8L
331684 -      ..
331685 -     At the top level, the only information is the name of the
331686 -     location. Presumably, there is a link to other information that
331687 -     would help to explain why a given location is good or bad for
331688 -     jobs, but the graph itself contains little or no pertinent
331689 -     information on the subject. ref DRT 1 SG9I
331691 -      ..
331692 -     At the second level of the hierarchy, the *only* information is
331693 -     the number of jobs. (Assuming that I am correctly interpreting the
331694 -     intent of the diagrams.) The individual bubbles would be useless
331695 -     for keeping track of jobs. They are already getting small and hard
331696 -     and select. And it would take different types of icons to present
331697 -     any useful information. ref DRT 1 00VX
331698 -
331700 -  ..
331701 - Graph Technologies Not Effective Collaboration Knowledge Management
331702 -
331703 -
331704 - Eric concludes...
331705 -
331706 -     Given these limitations, I don't see how graphing techologies
331707 -     apply at all to collaborative design/discussion tools or a
331708 -     knowledge base, given the huge volume of information such a tool
331709 -     needs to manage, the vast array of information types, and the
331710 -     exponentially exploding number of interconnects. ref DRT 1 RH4J
331712 -      ..
331713 -     Perhaps TheBrain has something that could provoke a change of
331714 -     mind. I can't say I've seen it (or recall what I saw, if I did).
331715 -     But as a simple example, how would any of the information
331716 -     contained in this message be captured in a graph? Were it done, in
331717 -     what way would such a graph be of use to anyone? ref DRT 1 00XZ
331719 -      ..
331720 -     I simply do not see graphing technology as useful in any
331721 -     substantive way in a knowledge-engineering context. It's GREAT for
331722 -     visualizing small systems, which makes it a wonderful tool for
331723 -     teaching. It gives people a mental model of the systems. But in
331724 -     actual use? I'm still inclined to pass, I'm afraid. ref DRT 1 TH5J
331725 -
331726 -
331728 -  ..
331729 - Representation Selected Elements Makes Complexity Comprehensible
331730 - Visualization Pictures Graphics Useful Often Essential
331731 - Pictures Worth 1,000 Words, Instantly Conveys Information
331732 -
331733 - Received ref DRT 2 0001 from John Deneen objecting to Eric's analysis,
331734 - reviewed above, ref SDS 0 UU8N, based on following comments....
331735 -
331736 -     STRONG DISAGREEMENT - "A picture is worth a thousand words."
331738 -  ..
331739 - Biological attraction of pictures was reviewed at...
331740 -
331741 -     Pictures convey instant impressions....940609, ref SDS 4 4933
331742 -     Pictures need intelligence.............940609, ref SDS 4 4671
331743 -     Pictures and Murphy's Law bumbling.....940609, ref SDS 4 567J
331744 -     Picture takes 1,000 words to explain...010312, ref SDS 16 Q56M
331746 -  ..
331747 - John continues...
331748 -
331749 -     Here's what some world-class scientists say about their need for
331750 -     visualization technology:, ref DRT 2 015N
331751 -        ..
331752 -        "We have all experienced the contrast between hearing (or
331753 -        reading) a description of a person's face, as opposed to seeing
331754 -        a picture of the same face.  The picture instantly conveys
331755 -        information to our minds which allows us to recognize a
331756 -        familiar individual. The narrative description may never convey
331757 -        enough information to arrive at the same conclusion. This
331758 -        ability of the human mind to rapidly perceive certain types of
331759 -        information, makes information visualization a useful and often
331760 -        necessary tool.
331762 -         ..
331763 -        Information Visualization is a highly efficient way for the
331764 -        mind to directly perceive data and discover knowledge and
331765 -        insight from it.
331767 -         ..
331768 -        Information Visualization is the direct visualization of a
331769 -        representation of selected features or elements of complex
331770 -        multi-dimensional data. Data that can be used to create a
331771 -        visualization includes text, image data, sound, voice, video -
331772 -        and of course, all kinds of numerical data. Our visual analysis
331773 -        systems also provide the tools to interact with the data that
331774 -        has been visualized so that users can explore, discover and
331775 -        learn. Users do not look at static images, but can subset the
331776 -        data, run queries, do time sequence studies and create
331777 -        categories and correlations of data type."
331778 -
331779 -            http://www.pnl.gov/infoviz/about.html
331781 -      ..
331782 -     Examples of important visualization tools:
331783 -
331784 -            http://www.pnl.gov/infoviz/technologies.html
331785 -
331786 -       [On 011102 Traction submitted examples using pictures for
331787 -       visualization of business meetings. ref SDS 23 3D4H
331788 -
331790 -  ..
331792 -  ..
331793 - Submitted ref DIT 1 0001 responding to John's letter on the important
331794 - role of pictures and graphs in human life, ref SDS 0 2X8K, and
331795 - objecting to Eric's analysis that these instruments are not effective
331796 - for knowledge management, also, per above. ref SDS 0 UU8N
331798 -  ..
331799 - Cite....
331800 -
331801 -      1.  Eric's analyis on 010312 that a picture takes 1000 words to
331802 -          explain. ref SDS 16 Q56M
331803 -          ..
331804 -      2.  Pictures are useful and often critical to daily
331805 -          affairs.
331807 -           ..
331808 -          The human mind interprets pictures by making associations
331809 -          with other pictures and drawing inferences of correlations,
331810 -          implications and nuance that apply, adjust and expand
331811 -          paradigms, rules, belief, past experience and huge a body of
331812 -          knowledge far beyond the confines of a particular picture,
331813 -          ref DIT 1 EY6O, as related in research reported in the record
331814 -          on 940609. ref SDS 4 V84K
331816 -           ..
331817 -      3.  Language is a more powerful, flexible window into the mind's
331818 -          interpretation than can be achieved by drawing a lot of
331819 -          pictures.  For example, it is very difficult to draw a
331820 -          picture of a concept, or to draw a picture that conveys the
331821 -          meaning of this paragraph, yet people readily draw useful
331822 -          meaning from the text. ref DIT 1 K77L This gives rise to a
331823 -          notion of an "alphabetic mind," reviewed on 991108.
331824 -          ref SDS 7 5628
331826 -           ..
331827 -      4.  Since speech is a powerful way to communicate at a much
331828 -          lower, i.e., refined, level of detail than is possible from
331829 -          reliance on pictures, an issue arises about how to manage
331830 -          what people say and do, since speech and action soon fade
331831 -          from memory. ref DIT 1 LU8M Henry van Eykan points out that
331832 -          people rely on remembering only the gist of things, about 5%
331833 -          of the actual record, reported on 000927. ref SDS 11 PO4F
331834 -
331836 -       ..
331837 -      No Perceived Threat or Gain, Laziness Encourages Poor Work
331838 -      People Care About Fast and Easy, Not Productive Useful
331839 -
331840 -
331841 -          We see some of this in our own work on the OHS/DKR, where
331842 -          people say things like, "...I remember so and so said a while
331843 -          back something about...., so I really feel thus and so..."
331844 -          The OHS/DKR project has not yet begun in earnest, so it is
331845 -          not critical that people do more than rely on remembering the
331846 -          gist of things. ref DIT 1 P39J  Since relying on the gist of
331847 -          things is fast and easy, there is no evident incentive to
331848 -          improve.  Like Jack Park explained for email on 010908, when
331849 -          there is no evident value at risk, no evident harm, laziness
331850 -          encourages taking action that is fast and easy; people don't
331851 -          care about being productive and useful, because people feel
331852 -          insulated from both harm of poor performance, and benefits of
331853 -          good performance. ref SDS 17 YF5O
331855 -           ..
331856 -          That is why manufacturers market computers for games and not
331857 -          usefulness, reported on 970603. ref SDS 6 4486
331859 -           ..
331860 -          However, on 900319 research shows memory is powerful part of
331861 -          human reasoning, indicating that poor memory significantly
331862 -          handicaps human intelligence. ref SDS 3 1323
331864 -           ..
331865 -          This research aligns with recent experience.  We have learned
331866 -          following tragic events on 010911 that the national security
331867 -          intelligence community has a lot of data and information, but
331868 -          cannot remember more than the gist of things. ref SDS 18 UP5K
331869 -
331871 -      ..
331872 -     Intelligence Demanded When People Get Hurt, Income Threatened
331873 -     Care About Safety and Income Increase When Threats Perceived
331874 -     Threats to Safety and Income Change What People Care About
331875 -
331876 -
331877 -          On 001207 we learned that too many people having too many
331878 -          problems relying on the gist of things, causes productivity,
331879 -          earnings and stock prices to suffer. ref SDS 15 V54M
331881 -           ..
331882 -          As with the OHS/DKR, so long as no one gets hurt, nobody
331883 -          cares if people use fast and easy methods that are not
331884 -          effective.
331886 -           ..
331887 -          However, it appears that many people do care about accurate
331888 -          memory when "the chips are on the line," i.e., accurate
331889 -          understanding of the actual record showing patterns of cause
331890 -          and effect impacts important values for guiding timely
331891 -          action. ref DIT 1 D44I  This suggests there is demand for
331892 -          "intelligence," which Eric proposed could be augmented, in a
331893 -          letter to the team on 000423. ref SDS 9 5096
331895 -           ..
331896 -          How then might this be accomplished?
331897 -
331898 -
331900 -      ..
331901 -     Intelligence Produces Knowledge to Be Prepared
331902 -     Organizational Memory Captured by Context of What People See Hear
331903 -
331904 -
331905 -          A textual explanation of what people say and do and hear and
331906 -          see, including any pictures, provides organizational memory.
331907 -          Adding "intelligence" to organizational memory creates
331908 -          history of cause and effect, explained in POIMS. ref OF 1
331909 -          2300 and defined on 890809, ref SDS 1 OU8O
331910 -
331911 -              [On 010924 organizational memory by itself is not very
331912 -              useful. ref SDS 22 L46F
331914 -           ..
331915 -          When "intelligence" is added that gives organizational memory
331916 -          context, this new resource provides the "knowledge" people
331917 -          need to be prepared for taking effective action, as explained
331918 -          on 950327. ref SDS 5 8822 Pictures are not conducive to this
331919 -          role, but still have a powerful role to preserve impressions
331920 -          and certain relationships, as in the design of a building or
331921 -          computer chip, etc. ref DIT 1 GS5L
331923 -           ..
331924 -          In sum, pictures and text alone are not enough.  We need to
331925 -          follow Eric's lead and develop useful intelligence, defined
331926 -          in POIMS. ref OF 1 0367  Jack, Eric, Eugene, Lee and others
331927 -          seem to be working along these lines.  Yesterday, on 010913
331928 -          Eric reported that Eugene will have an announcement soon.
331929 -          ref SDS 20 A75O  The rest of us need to be developing the
331930 -          culture and work practices for deployment, as Eugene
331931 -          suggested on 001126. ref SDS 14 BY4K  Doug has been writing
331932 -          about this for years, ref SDS 8 3696, so there is no time
331933 -          like the present to get started producing useful
331934 -          intelligence. ref DIT 1 6W6L
331935 -
331936 -
331937 -
331938 -
331939 -
331940 -
331941 -
331942 -
331943 -
331944 -
331945 -
331946 -
331947 -
331948 -
331949 -
331950 -
331951 -
331952 -
3320 -
Distribution. . . . See "CONTACTS"