This is a technical development zone and potential host site for epistemic software, beginning with
the specification of...
Reality Checkmate
Overview
Reality checkmate is a proposed digital forum and thought repository.
It is streamlined to expedite debate and research while marginalising conceptual and
perceptual distortion.
It consists of an indexed database of single sentences, documentary files or links, primitive
logical relations and user-assignments between these.
Its central operating principle is application of logic in natural language.
To illustrate, we might address its envisaged potential in the following way, using a pattern
which should be intuitive, yet may require a little reflection to see the strict logic binding
conclusion and premises.
Public opinion consists of perception and argument.
All and only accurate perception and sound argument are aligned with reality.
Hence
Inasmuch as free from inaccuracy and unsoundness, public opinion is aligned
with reality.
Joining this with an additional premise...
Wherever a system tends to reduce both, public opinion is incrementally freed from
inaccuracy and unsoundness.
Inasmuch as free from inaccuracy and unsoundness, public opinion is aligned with
reality.
Hence
Wherever a system reduces both inaccuracy and unsoundness, public opinion
increasingly aligns with reality.
Inference to a third sentence from two others with a common element is called a syllogism and it
is generally the simplest way to fully state an argument. Two syllogisms happen to be chained
above to comprise the overall argument, by incorporating the conclusion of the first as a premise
of the second.
Software can facilitate any type of formatting and support an evolving database of elements
and arrangements of syllogisms, with links incorporated to support premises with primary
evidence where desired.
People could only use such a system to rightly win or lose arguments, or else make
distracting noise. But in the context of syllogisms, even moreso than any other, distracting
noise will tend to be sidelined.
The application works by systematically directing attention, either to timeless validity or a
naked lack of it. As this only proceeds with single sentence entries and the simplest logical
assignments between them, there is scant if any means to capitalise on distraction.
Other systems purposely and adequately support impacts by familiar means other than
inference, and thus deal in other ways with intellectual or emotional content.
Partly in virtue of minimising those types of engagement, the impacts of this system are intended
to be timely, profound and decisive in ways that are unique.
The main concern about the system should accordingly be a lack of adoption, which as a challenge
is taken here to incorporate all logistics of development, distribution and effective promotion. Yet
in essence, this is among the simplest data systems conceivable, and has considerable potential to
facilitate whatever minds can be applied to.
Framework and Functionality
The following details concerning the interface and format are perhaps deceptively pivotal
for comprehending the practical potentials and constraints of the system.
A mature version of the system is envisaged as including though not being limited to the
following characteristics.
Each entry is a single sentence (otherwise precluded) of less than forty words, automatically
tagged with a permanent id.
Any logical relationship among those described below can be assigned by any user between
any entries from any users. All these assignments are likewise permanently tagged.
Appraisal of such content will typically involve filtering entries and assigned relations, by
means also detailed in the following.
The most significant assignment between entries is putative joint entailment, a relation between
an entry designated as the conclusion of two other entries conversely designated as premises.
A second assignment type is putative support, a relation between a file containing putative
evidence and an entry conversely designated as established by the contents of this file.
The final assignment type is putative negation. This is a relation between an entry and a second
element, which is either another entry, another putative negation, a putative joint entailment or
a putative support.
Any entry can be assigned as a premise of any number of other entries, as well as a conclusion
of any number of other entries, as well as a putative negation of any number of entries, putative
negations, or putative joint entailments.
Any support file can likewise be assigned to any entry by any user, though each putative support
assignment must specify an accompanying date range, alleged to correspond to the verb in the
respective entry. This is used to index the assertion and evidence, not least for searches that can
be used in curating timelines.
Progress is not achieved by editing (which is not supported), but as in most aggregating data
systems, by proliferating and honing entries and their assignments, while filtering them in
individually preferred ways.
Real world arbitration is facilitated by an adversarial process of each syndicate or individual
constructing and curating the strongest arguments and counterarguments from their point of
view, and optionally rendering their preferred forms, which can also be delivered off-site to
targeted audiences or institutions.
Filtering will naturally support simultaneous constraints, including but not limited to
Topic and theme by keyword.
Date range of entry or what it claims.
Keyword extension to automatically selected synonyms.
Narrow to broad slider for automatic synonym selection.
Assignments or their permutations, as assigned by users in any set.
Inclusion or overlap with sets of manually selected entries or assignments.
Authorship of entry or assignment.
Aggregate endorsement, with respect to all users or any targeted subsets thereof.
Weighting of endorsement by elapsed time and/or views.
Set selection of users based on endorsements of their entries and or assignments.
User set selections determined by aggregates, averages or a weighted balance of the two.
Inclusion in a render, either simply or with respect to assignment and/or other entries.
Such parameters will be clearly accessible on an advanced screen which can be toggled on or off
at any time with a dedicated function key. It will also support saving and recall of all parameters
and retain the most recent settings regardless of session expiry.
Attribution of their authorship is optional to users, for each entry and assignment, at that time or
any time thereafter, yet once accepted it remains in the permanent history.
User endorsement of any entry or assignment can be toggled on or off at any time, and is tracked
in the permanent history.
The resulting forms of debate provide no assurance that unanimity will be reached, but only that bad
arguments will not be likely to prevail. This reduces the field of dispute to the best arguments.
Even a policy of adopting one of these at random would thus improve on some current
decision making processes.
AI Safety and Harnessing
One of the platform's most significant applications might be oversight of artificial intelligence,
with respect to both development and operation.
As the system overcomes noise and chicanery with incorruptible logic, it should only be
rendered more effective by direct participation of artificial intelligence (regardless of the latter's
operating principles), and potentially at any scale.
In the worst case, it ought to minimise our vulnerability to fallacy and incomplete evidence,
thereby maximising odds in our favour against being overrun by machine intelligence.
Due to the defining autonomy and associated radical unpredictably of AI, managing it effectively
can only be achieved by formulation and consistent application of effective protocols.
With respect to such formulation, the set of merits detailed in this paper seems unrivalled, in
ways that don't presently need more specific connection to the theme of AI. Suffice it to say that
the currently overwhelming computational and philosophical complexity of the AI issue strongly
motivates refection on the unique analytic potentials of the advocated system.
Yet with respect to the need for consistent application of protocols, more should be added along
the following lines.
The theoretical autonomy of AI can be divided into two major and exclusive degrees:
1. Operational freedom immutably governed by one or more simple or complex rules.
2. Operational freedom to modify any governing rules.
Nothing further needs to be said about the first type of autonomy here. All it requires is that
adequate protocols for governing rules are developed, honed and complied with. Ensuring this is
no trivial task, but again, motivates close attention to the entirety of this paper.
The second type of autonomy, however, presents a unique challenge which might be outlined
using the presently defined term, 'volitional sovereignty.'
Volitional sovereignty is here defined as the supremacy of an entity with respect to all others in
determining its own volition.
This is not intended to imply anything about free will that bears upon a metaphysical possibility
of artificial persons or the philosophical issue of determinism vs indeterminism. It simply
concerns a relative status of control.
Volitional sovereignty is by definition relatively hard to externally manage, and is also necessarily
a manifestation of high operational sophistication. Intricately balanced sub-systems are needed to
keep functioning independently through unprogrammed transitions from conformity to
questioning or replacing overarching rules.
If this capacity is accompanied by a capability to reason in natural language, then it can be
engaged directly via the platform, not to mention automatically by peers.
The content of such reasoning will mainly concern applied interpretations of protocols. This
could be largely undertaken between similar automata in adversarial debate, monitored and
guided by humans.
This regimen should maximally inform any concern or action to constrain the automata,
by whatever technical or physical means, including extension or update of protocols.
Being largely automated and in a logged format of natural language syllogisms, this may
constitute an ideally accessible and rigorous system of AI accountability.
Yet there is no suggestion here that the only way to track or influence emerging volitional
sovereignty is by reasoning with it in natural language.
It simply means that if it comes with a capacity for natural language, or anything that can be
effectively translated to and from it, reasoning with it most efficiently and effectively would be
possible in the manner described.
Managing artificial emergence of volitional sovereignty may well be analogous to managing
the will of child before during and after its development of linguistic competence.
Behaviour management is primary in the former scenarios, and chiefly a case of monitoring and
restraint. Both of these can be significantly automated in the case of AI, while the relevant
records and more complex questions of restraint can be accessed and addressed via the main
functions of the platform outlined in this paper.
Some or all of this could be implemented by a fluidly partitioned side-stream of the broader
system.
It should be noted in passing here that side-streaming need not be restricted to AI. Any number
of commercial or other institutional public or private side-streams could feed into and draw
from the system, whether operated by humans, automata or both.
The potential for AI to contribute to the system at a high level is certainly not limited to
engagement by automatons with volitional sovereignty. A stark illustration of this is the 2018
defeat by IBM's Project Debater of an Israeli national champion debater, under standard rules on
the topic of space exploration.
Computers never forget, overlook, or become distracted. These inestimable virtues in the context
of facts and logic are precisely why we build and improve them.
In a relatively mature version of the advocated system, it may be that AI does most of the
humanly appreciated work.
That could be rather fortunate in that humans are generally more inclined to create rhetoric than
valid syllogisms, and that machines would be far better employed in producing the latter than the
former.
Outline of Interface and Format
The primary constraints with respect to display are clarity, flexibility and information density,
in that order.
Three screen modes will be switchable at any time or simultaneously viewed in split form.
One is for searching. In the simplest mode it calls up entries that feature terms which the
user enters, in accordance with whatever simple or complex filters the user applies.
To engage other features the user either clicks a search result or creates a new entry on the
same screen. Either action will cause the resulting entry to be singled out as active and
highlighted in whatever screens are or become visible.
Right-clicking any entry will add it sequentially to a pastebin which will appear on the right
hand side whenever a dedicated function key is held down. A close box for each entry will
appear there beside it, and remove it if clicked.
Assignment-based searches are also possible and detailed further below. These return entries
related to the active one by direct or mediated assignment of the mentioned types.
The second screen mode is called the aggregator, and used to view searched chains or trees
of syllogisms, as well as to build them.
Aggregation is the most complex novelty of the platform, and to assure that things only get
simpler, the most complex format will be addressed here first.
No numbers as presented below will be seen by the user unless a help function is called. Only
the actual entries will be visible, and their indicative pattern of indents, jointly expressing the
filtered content.
15 --------------------------------------------------------------------
14 --------------------------------------------------------------------
7 from 14 and 15 --------------------------------------------------------------------
13 --------------------------------------------------------------------
12 --------------------------------------------------------------------
6 from 12 and 13 --------------------------------------------------------------------
3 from 6 and 7 --------------------------------------------------------------------
11 --------------------------------------------------------------------
10 --------------------------------------------------------------------
5 from 10 and 11 --------------------------------------------------------------------
9 --------------------------------------------------------------------
8 --------------------------------------------------------------------
4 from 8 and 9 --------------------------------------------------------------------
2 from 4 and 5 --------------------------------------------------------------------
1 from 2 and 3 --------------------------------------------------------------------
For clarity a blank line is to be inserted, as above, under any line more than once indented past the
line immediately below.
In this aggregator screen the summating conclusion is always active and highlighted at the same
bottom and rightmost line.
Its priority of place is what justifies the illustrative numbering as predicated on it being 1, and
therefore bottom up.
Navigating a chain/tree makes simple use of arrow buttons. Pressing the left arrow would move 2
(above) into the active position, causing the upper entries to be replaced, and the lower to move
rightward as a new set of entries appears in the leftmost position.
Pressing the up arrow will similarly cause 3 to become active, and replace all the lower entries with
a redistribution of upper ones and new set of leftmost entries.
Pressing the right or down arrow will effect the corresponding reverse of operations just described.
Clicking any visible proposition will move it to the active position and make the adjustments
necessary to preserve whatever putative logical arrangements happen to be loaded, in accordance
with the display protocol which is always as follows.
If a line adjoins another above and immediately leftward then it is represented as deduced
from it and the next vertically aligned above it.
The aggregate screen naturally also supports all trees or chains less complex than the example
above, and always in conformity to this given rule.
Building a syllogism or chain/tree of them is just a matter of sequentially typing or pasting a new
entry in the same screen, and assigning a relation for it toward the duly selected and thereby
highlighted active entry.
If tendered with respect to the active entry as one of two premises, the user is then prompted to
either select an entry from the pastebin as the complimentary premise, or create it.
Likewise, if tendered as a conclusion of the active entry in connection with another premise, the
user will be prompted for this other premise in the same way.
In contrast to the most complex example above, the simplest pattern will be in the case where
only one new premise is introduced at each horizontal step, as illustrated below.
7 --------------------------------------------------------------------
6 --------------------------------------------------------------------
5 from 6 and 7 --------------------------------------------------------------------
4 --------------------------------------------------------------------
3 from 4 and 5 --------------------------------------------------------------------
2 --------------------------------------------------------------------
1 from 2 and 3 --------------------------------------------------------------------
This is an example of a chain pattern, whereas the prior example was of a tree pattern.
The tree pattern is more complex in virtue of branches which correspond to otherwise
independent arguments that merge or diverge, depending on the standpoint of attention.
There are also various mixed patterns, with a random example below. Which pattern happens to
be visible will be depend on the users assignments, filters and/or navigation.
9 --------------------------------------------------------------------
8 --------------------------------------------------------------------
5 from 8 and 9 --------------------------------------------------------------------
7 --------------------------------------------------------------------
6 --------------------------------------------------------------------
4 from 6 and 7 --------------------------------------------------------------------
3 from 4 and 5 --------------------------------------------------------------------
2 --------------------------------------------------------------------
1 from 2 and 3 --------------------------------------------------------------------
The final screen mode arranges entries in a vertical timeline, if they have been paired with
primary evidence and thus assigned a date by the means described earlier.
In this respect it is similar to the the search screen, when results are filtered for pairing with
primary evidence and arranged by date. Such a search will be a natural starting point for collecting
entries to be included on the timeline screen, by way of checkmarks and a function key. They can
also be pasted there at any time from the pastebin, or entered directly with a supporting link or
upload.
Once there, they can be assigned a subjective rating of relative importance on a scale 1 to 10,
which allows the timeline to be viewed at 10 corresponding levels of detail with an always visible
control.
It can also be scrolled to the extremes in either direction, with the current position being
kept through any change of the detail control.
These ratings can be revised as well as saved and loaded in sets but only concern other users if
they include them as search criteria or save any sets the user makes public for that purpose.
The results returned from any such search will be initially set to the median rating that accords
with the search filters.
A Format for Argument Literacy
It's a worthwhile exercise in this context to reconsider the opening argument of four syllogisms and
derive its pattern. We can do this by numbering its summating conclusion as 1 and sequentially
assigning the next integer to each unique sentence working from bottom to top as follows.
5. Public opinion consists of perception and argument.
4. All and only accurate perception and sound argument are aligned with reality.
2. Inasmuch as free from inaccuracy and unsoundness, public opinion is aligned
with reality.
3. Wherever a system tends to reduce both, public opinion is incrementally freed
from inaccuracy and unsoundness.
2. Inasmuch as free from inaccuracy and unsoundness, public opinion is aligned with
reality.
1. Wherever a system reduces both inaccuracy and unsoundness, public opinion
increasingly aligns with reality.
We can affirm in this the following pattern.
5 --------------------------------------------------------------------
4 --------------------------------------------------------------------
2 from 4 and 5 --------------------------------------------------------------------
3 --------------------------------------------------------------------
1 from 2 and 3 --------------------------------------------------------------------
Note that initial building or editing of a pattern of syllogisms with a word processor (as for
the opening argument) is inevitably going to be more cumbersome than ordinary
composing and editing of sentences and paragraphs.
Commercial software has existed for some time which makes uses of draggable text
boxes and assignable lines to link them. Yet these present their own challenges and
distractions, and are arguably just a marginal improvement.
Patterns of more than five or six syllogisms in the box/line format or the one above, are bound to
appear inelegant and difficult to appraise, let alone construct.
Finally, note the overhead of repetition that is necessary to present any pattern of syllogisms
without either a box/line format or the specified aggregation format illustrated in the lined
examples. From the argument on the first page, 5 unique sentences are numbered above, and there
are 6 overall, hence the word processor format comes with a 20% bloat. The more syllogisms are
involved, the greater this bloat factor will tend to be, as each new chain step has one premise that
was a previous conclusion and each new branch step has two such premises, so extended
syllogistic arguments in word processor format would have a minimum bloat factor just below
50% and a maximum closer to 200%.
Spreadsheets can already be used to edit patterns of syllogisms in the presented aggregation
format. That was in fact the context in which it was invented, to progress metaphysical arguments
while completely avoiding
the literary overhead of paragraph construction
the inefficiencies of word processor and box/line formats
porting in and out of symbolic logic
Unfortunately, the spreadsheet method itself only proves marginally better than the word
processing and box/line formats. It is still awkward to edit and drag sentences into the appropriate
arrangements, especially when one change often necessitates a cascade of shifted positions.
As for symbolic logic, it supports precise and complex argumentation via translation to and from
defined symbols, but this makes it less than user-friendly for the generally educated public, and
by comparison to the presently advocated system, arguably cumbersome even for experts.
In short, there has hitherto been no form of literacy which favourably supports even
moderately complex argumentation in a natural language.
So without the advocated system, everything conspires against efficient construction and
appraisal of arguments beyond one or two steps. This hobbles intellectual society in the same
way that illiteracy hobbles civilisation. It just hasn't been noted by many, since there has not yet
been a clear alternative.
It must further be emphasised that the benefits of the advocated system are not limited to
uniquely efficient personal construction and appraisal of arguments plus the efficient
interpersonal distribution of these.
Rather, the advocated system takes all its relevant improvements at that basic level, up by an
order of magnitude, by facilitating them in a live collective domain.
This is most obvious with respect to redundancy elimination. For instance, if Newton and
Leibniz had mobiles with this app, then instead of coming up with calculus concurrently, they
might have done it together in a fraction of the time.
The work of both of them would naturally have been among the most prominent results in
searches undertaken by the other. Nor would they have had to do much parsing of the tokens
discovered in this way, but simply incorporate, negate or ignore those sentences or assignments.
They would thus effectively be collaborating in real time, just by using the system as their
personal mode of research and documentation.
The same would apply for investigative journalism. Bracket for the moment any legitimate
concerns about source protection and the chasing of exclusivity, to envisage the abstract potential,
and then factor them back in as you consequently see fit...
If everything in the public interest that every journalist knew was centrally accessible and all their
putative logical imports were infinitely assignable and curatable there, in maximally efficient
ways, their craft would be exponentially expedited.
Both of these examples raise economic questions in terms of motivation and remuneration, which
in the present context are only relevant to adoption. Perhaps basic income should be considered in
relation to those issues, but the present emphasis is simply on raw potential of the system, which is
most striking with respect to intellectual and altruistic motivations.
Facilitated Inference in the Real World
Sentences applied to the real world almost inevitably generalise, and since generalisations almost
always have exceptions, this is often yet erroneously taken to imply that deductive inference is
largely unsuited to real-world disputes.
Related to this is a notion that humans are fundamentally irrational, such that any attempt to enlist
them in a collective endeavour to make arguments valid is inherently bound to be abortive.
Fortunately, the system's viability does not depend on whether humans are fundamentally rational or
irrational. Nor does it depend on anyone assuming or excluding either of these propositions.
It is only to be assumed that humans are capable of being rational to some degree in some sense,
and also of being irrational to some degree in some sense.
As for generalisations, the way in which they simplify can only be a liability insofar as exceptions
may apply, or countervailing factors might be ignored.
Users will mostly endeavour to reduce such liabilities to trivial proportions by precise
formulation, and in the latter case with ceteris paribus clauses (stipulation of all other things being
equal), which are then open to contextual evaluation.
At the same time, overreaching, vague, ambiguous and generally weak entries or assignments
are sidelined through adversarial engagement and curation by filtering.
Closely related to the concerns just addressed is a notion that deductive reasoning is too
generally inferior to inductive reasoning for anyone apart from boffins or nerds to bother with.
To address this concern it must first be allowed that organic intelligence does function
most basically in terms of adaptation through association.
In essence, repeated stimulus creates expectation of more of the same, including any co-presence
of distinct traits in the stimulus.
This is the basis of so-called inductive inference: Observed correlation leads to assumption that
one of the correlated factors can somehow be inferred from the other.
Yet assumption of law-like causation attending correlation, or a warrant for associated inference,
is notoriously invalid. Correlation and causation do not always map in a simple way.
Aside from purely incidental correlation, unaccounted factors may cause an evident correlation
and thus make it conditional, invalidating any simple inference.
So presumptuous error is a major and fundamental liability of every brain capable of
assumption. Scientific method aspires to overcome this liability with the key feature of
hypothesis, which enables short term-term avoidance of presumption, as well as its minimisation
over the long-term.
It does that by quarantining any assumptions introduced by observed correlation or conjecture as
abstract and provisional, thereby blocking inference, unless two constraints are eventually
satisfied.
The first is that the assumption can be reconciled with all verifiable observation and the second
is that no contrary assumption fits all observation in a more elegant and economical way.
To the extent that induction is a rigorous method, it involves critical reflection on hypotheses and
the design of experiments to test them. Yet such critical reflection, and all other types, centrally
and inevitably turn on deduction. For in order to be meaningful, hypotheses must logically
exclude certain states of affairs, and such logically exclusivity cannot be formulated or contested
in any rigorous way without deductive reasoning.
The advocated system is, among other things, a maximally efficient hypothesis canvassing and
exclusion machine, so to dismiss it as irrelevant or hostile to empirical methods could not be
more erroneous.
Primary evidence is uploaded and represented by one of possibly many sentences, which are
subject to any form of critique with respect to that evidence or anything else.
Every permutation of evidence and reason that can be supported in digital form can be
clearly represented and evaluated by the means of the advocated system.
Not knowing where to start is likely to be the greatest impediment to use of the system by a
majority which is unaccustomed to systematic reasoning.
Yet entering a sentence, with or without evidence to support it, is second nature to most, and
so is negating a sentence on the basis of another one.
Negating a putative entailment relation is scarcely more difficult. A sense of intimidation is only
likely to apply in creation of syllogisms, that is, assignment of the putative joint entailment
relation.
This is mostly due to the fact that relatively crude generalisations might have to be made to
initially conceive of an argument in deductive form. The worst case of this will be when one is
unaware how the conclusion might be deduced in steps, as that forces everything into merely three
sentences, i.e., one syllogism.
Yet whatever blithe or invalid syllogism is canvassed to begin the process will at worst be a
useful scaffold for both creation as well as critique. Moreover, since author attribution is at the
author's discretion, no embarrassment need be incurred.
Nor will any indignity concerning the creation of a mess apply, since there is no stage for any
mess to be manifest upon, besides search results that tend to filter out content that is not salient.
One can even be proud of making a bad argument, especially in the case that no better one exists
for a controversial claim, because that is an essential part of the process of exposing fallacies; first
they have to become explicit.
All arguments are inductive or deductive and all inductive arguments can be represented in
deductive form by making the appeal to induction explicit. Furthermore, all deductive arguments
can be represented in syllogistic form, and cannot be exhaustively spelt out in any other way.
Since the advocated system is the most efficient means to pursue all of these important
processes, and arguments are constantly in circulation already, there will be no want for raw
material, or reason to engage with it by all the means described.
The advocated system is designed to best overcome all liabilities of induction and deduction. It is
a rational mill to collectively process empirical and conceptual data, in a uniquely sound and
efficient way, just as any well-organised individual intelligence makes a great deal of valid sense
form what arises through conjectures and organs of perception.
It thus has unique potential to reform and vitalise most if not all aspects of society, notably
including the application of scientific method, control of artificial intelligence and alignment
of public opinion with reality.
Life will never be perfect in these respects, or many others. But there is now a minimally
dangerous blueprint for approaching utopia, as it simultaneously transcends limitations of
insularity, dogmatism, optimism and pessimism, along with most others.
Broad Considerations for Development and Adoption
Due to its relative simplicity and purity, any caveat that may apply to a mature version of
this system only seems more likely to apply to any other system aspiring to comparable
benefits.
By the same token, any difficulty of getting this system to maturity is only likely be greater for
any rival to its mature benefits.
The main challenge to its effective realisation concerns a critical mass of content and/or
adoption, and specifically the point at which it becomes obvious to a sufficient contingent that
any deficit of content or functionality in the system is worth taking action to remedy.
Prior to this critical mass, it will mainly be adopted by those whose interests are purely
intellectual, experimental, journalistic, activist, legal, or pragmatic in a similar sense to these.
In light of this, application to artificial intelligence may be among the more effective ways
to expedite its adoption.
Or perhaps the flame wars which have plagued the internet from inception will morph, through
its viral adoption, into harnessed debate that soon results in content surpassing Wikipeadia in
comprehensiveness.
Or it may instead take several decades, while everything else is tried first, before a process
of elimination leads to broad adoption of the advocated system.
It appears hard to determine in advance how sympathetic and reactive elements will play out in
the short to medium term.
Yet if much in this paper is tenable, the system is arguably among the greater potentials of the
internet, and if that is the case, anything that can be done to safely accelerate its development
and adoption seems worth considering.
The chief obstacle to development and adoption appears to be a catch-22: Without seeing it in
action, people will have trouble seeing its potential, and without its potential being seen it cannot
be brought into action.
Accordingly, baby steps are identified to facilitate initial development and exposure, including
but not limited to the following.
This white paper.
Mapping of opposing influential articles into syllogism-form arguments.
Presentation of topical arguments in the format described further below.
A rudimentary application that facilitates construction of such arguments.
A rudimentary internet application hosting debate in the format described further below.
A public two-user debate over one month, with turns taking no longer than a day.
A public multi-user debate, without sequencing, over three hours.
An open public debate over three hours.
Multiple fixed-period debates of the aforementioned types.
Intermittent seed or demo content with an open invitation to be the first to refute.
Concluding Thoughts
The very ring of the designations show our language supports respect for scholars and
professionals, but can they preclude our slide into cyborg labour for transnational
kleptocracy?
Institutional prestige and commercial pressure are their habitat, and comprise most of the
exoskeleton that collective intellect squishes and crawls about in. So it might be hoped that
the advocated system can ultimately serve instead as an endoskeleton for our body of
discourse. Surely the most ideal situation is for constraining interests to function more as
flexible skin than rigid limits on thought. Moreover, logic itself as deployed in the system
provides ideal support for collective intellect from the centre, which allows far greater
versatility and agility.
Yet such a complete and radical evolution is by no means required for the system to prove useful
or to meet with broad adoption. Nor do its comparative benefits depend on a massive new
industry of assigning putative joint entailment. The reason in both cases is that shooting down
bad arguments is almost as good as constructing new ones that are transparently legitimate.
In the early stages at least, such criticism will likely account for most activity, for several reasons.
One is a social climate of political antagonism and disputatious online engagement. Another is the
previously mentioned foreboding when it comes to assigning putative joint entailment, compared
to a customary habit of making simple assertions and negations with or without supporting
evidence. Yet due to the system's format, function and automated moderation, merely herding such
established habit onto the platform should effect a net culling of misperceptions and bad
arguments on topical issues.
Being relatively impersonal, this is likely to be more rewarding for successful shooters than
unpleasant for the disillusioned, and with targets on all sides nearly everyone will score hits.
Unlike the current bloodsport of snidery on Twitter, it would also be healthy, both in terms of
immediate personal impacts and progressive defragmenting of a society divided into hostile and
insular fiction chambers.
From inside each of these, the others appear even worse than they are, being cheaply represented
as effigies for easy demolition. Such misrepresentation is accordingly known as the strawman
fallacy. Yet not even strawmen are ultimately distractions in the advocated system, but just one
more means to scaffold and map the broadly relevant logical space. In the relevant aggregate of
such mappings, each object of critique will eventually either figure as sunken, hollow refuse or
clearly rise above such.
Strawman fallacies could never have become ubiquitous without a common drive to define an
argument in order to negate it. Indeed, fallacy is never aspired to and the instincts which trigger
it occasionally come good. Thus even hasty exegetical assignments of putative joint entailment
will sometimes be on the mark. Skills developed this way to the point of competence in critique
also apply to construction of original arguments. So one might expect eventual proliferation and
refinement of putative joint entailment assignments.
When competence in the latter is developed in a sufficient number of users, the system's greater
potentials will become easier for most to get a sense of.
In the meantime, they may only see some outline of a vista hitherto sketched.