Online scientific discourse is broken and it can be fixed

Has science taken a wrong turn? If so, what corrections are needed? Chronicles of scientific misbehavior. The role of heretic-pioneers and forbidden questions in the sciences. Is peer review working? The perverse "consensus of leading scientists." Good public relations versus good science.

Moderators: MGmirkin, bboyer

Re: Online scientific discourse is broken and it can be fixe

Unread postby Plasmatic » Fri Dec 06, 2013 10:05 pm

Lloyd, I liked that truth mapping vid. It seems there's already tons of tools available.
"Logic is the art of non-contradictory identification"......" I am therefore Ill think"
Ayn Rand
"It is the mark of an educated mind to be able to entertain a thought without accepting it."
Aristotle
Plasmatic
 
Posts: 800
Joined: Thu Mar 13, 2008 11:14 pm

Re: Online scientific discourse is broken and it can be fixe

Unread postby CharlesChandler » Fri Dec 06, 2013 11:12 pm

I found truth mapping to be feature-poor. For example, I can only add a critique, not a rebuttal. I guess that the thread starter can add premises and rebuttals, and everybody else can only add critiques. That makes for a me-against-you debate. But I don't see how it supports a collaborative environment. We need the ability for anybody to make contributions wherever they are relevant.
Give a man a fish and you feed him for a day. Teach a man to fish and he'll spend the rest of the day sitting in a small boat, drinking beer and telling dirty jokes.

Volcanoes
Astrophysics wants its physics back.
The Electromagnetic Nature of Tornadic Supercell Thunderstorms
User avatar
CharlesChandler
 
Posts: 1790
Joined: Tue Aug 19, 2008 6:25 am
Location: Baltimore, MD, USA

Re: Online scientific discourse is broken and it can be fixe

Unread postby Lloyd » Sat Dec 07, 2013 8:41 am

CC: I found truth mapping to be feature-poor. For example, I can only add a critique, not a rebuttal. I guess that the thread starter can add premises and rebuttals, and everybody else can only add critiques. That makes for a me-against-you debate. But I don't see how it supports a collaborative environment. We need the ability for anybody to make contributions wherever they are relevant.

A group could arrange to use the truthmapping site collaboratively, although it would perhaps be better to have a similar template with more versatility, where the owner/author could allow others to act as co-owners etc.

Charles, can a similar but more versatile template be constructed on your site? If so, I'm ready to try it out there.

Deliberatorium does allow owners to give other users co-owner privileges, but the template there seems less suitable for scientific papers, although it might be flexible enough to be made so.
Lloyd
 
Posts: 4317
Joined: Fri Apr 04, 2008 2:54 pm

Re: Online scientific discourse is broken and it can be fixe

Unread postby CharlesChandler » Sat Dec 07, 2013 2:23 pm

Lloyd wrote:Charles, can a similar but more versatile template be constructed on your site?

Here's an example of the format that seems to work the best so far. It starts with a brief title of the issue (such as "Black-body Radiation", or whatever it is). Under that are two items: Observations, and Explanations. The Observations are the data, stated in a theory-independent way. Then come the Explanations, starting with the initials of the person or group doing the explaining, where there can be as many as there are people willing to venture opinions, but there shouldn't be any duplicates (i.e., only one item per distinctly different explanation). Then, people get to add sub-points below each explanation, for criticisms, rebuttals, etc. There again, the rule is just that there shouldn't be any duplicates -- each idea should appear only once, exactly where it belongs in the outline. The main advantage to this, and any other structured argument, is that it eliminates all of the tail-chasing, flooding, and misdirection that are so problematic in journal-style threads. I think that we should try using it just like this for a little while, and see what problems come up, and then, see what additional structure needs to be applied to keep the debates on track.

Black-body Radiation
  • Observations
    • The Sun emits 3.8 x 1026 watts, in the form of 5525 K black-body radiation, (4600 K on limb, 6400 K normal to surface), with some absorption and emission lines, and is typical of main sequence stars.
  • Explanations
    • Std: fusion in core generates gamma rays that propagate through radiative zone, getting redshifted by a wide variety of non-BB processes
      • CC: how is the model pressure in the core sufficient to overcome the Coulomb barrier in the core?
      • CC: convection has been found to be insufficient to carry enough heat to the surface, so how does the sparse plasma conduct the heat?
      • CC: what are the chances of a combination of non-BB processes producing a BB curve?
    • WT: Electric current through Sun causes ohmic heating.
      • CC: current from where, and to where?
    • MM: .7 Mm deep neon layer is 4000 K hotter than underlying silicon, so most of the BB radiation comes from the neon layer
      • CC: what about limb darkening — could a layer only .7 Mm deep vary from 4600 K to 6400 K, with granules redistributing the heat every 20 minutes?
    • CC: atomic oscillation in supercritical fluid at a depth of 4~20 Mm, due to ohmic heating from electric current, where electrons are emitted by a negative layer at a depth of 20~125 Mm, attracted to the positive heliosphere
Give a man a fish and you feed him for a day. Teach a man to fish and he'll spend the rest of the day sitting in a small boat, drinking beer and telling dirty jokes.

Volcanoes
Astrophysics wants its physics back.
The Electromagnetic Nature of Tornadic Supercell Thunderstorms
User avatar
CharlesChandler
 
Posts: 1790
Joined: Tue Aug 19, 2008 6:25 am
Location: Baltimore, MD, USA

Re: Online scientific discourse is broken and it can be fixe

Unread postby Lloyd » Sat Dec 07, 2013 7:59 pm

Charles, are you wanting to start that here, or on your site, or what? I'm willing to participate, if you decide where to start. But I'd like to see a feature like on truthmapping for critiques and rebuttals. Comparing a bunch of disparate theories seems too overwhelming.

Something I like in your theory is showing that there is only one alternative that's workable. E.g., you explained that the electric force is the only one that can account for the smooth, distinct limb of the Sun. Then you explained that several double layers can account for the basic solar features, but that only one arrangement of double layers can work to hold the photosphere and solar atmosphere etc in place. It seems to me that that is the Occam's razor approach, which has the advantage of simplicity, cutting off all the stubble theories.

I'd like to see what Zane, Plas and others might think too about all this.

On the NIAMI board I referenced my new truthmapping paper for all of science, maybe, but starting with physics and astrophysics. Don't you all think it might be helpful to have a logic tree for all of science? I think what I started there
http://thunderbolts.info/forum/phpBB3/viewtopic.php?f=10&t=14757&p=90684#p90684 might be expandable to include all of science logically.
Lloyd
 
Posts: 4317
Joined: Fri Apr 04, 2008 2:54 pm

Re: Online scientific discourse is broken and it can be fixe

Unread postby CharlesChandler » Sat Dec 07, 2013 9:48 pm

Lloyd wrote:Charles, are you wanting to start that here, or on your site, or what?

We can't do it here, because we need to use software that supports creating/editing structured discussions, which would mean my site (or TruthMapping, Deliberatorium, DebateGraph, TopicCentral, etc.). I prefer my site, because it's more flexible, and I don't think that we're sure what we want yet. ;)

Lloyd wrote:But I'd like to see a feature like on truthmapping for critiques and rebuttals.

I'm not sure exactly which feature about TruthMapping you're after here. Like I said before, I found TruthMapping to be too limited in features for what we're trying to do. Nevertheless, if it has an advantage, let's see if we can instantiate that in QDL. Is it that you want the premises, critiques, and rebuttals clearly labeled (like issues, ideas, pros, & cons in Deliberatorium)? I was thinking of making the icons in QDL user-selectable. So you could go with the standard gear icon, or you could select a different icon, to designate what kind of item it is. (BTW, who trashed all of the stuff that I entered into the Deliberatorium?)

Lloyd wrote:Comparing a bunch of disparate theories seems too overwhelming.

I'm thinking that there should be a section for "New Debates", that would allow people to just willy-nilly start a new debate on whatever interested them. With the cross-referencing in QDL, this debate could be left in that folder, but also linked into the master outline at the appropriate place. So people wouldn't have to get overwhelmed by seeing the whole structure when just discussing one particular thing. But anybody looking at where that discussion fits into the master outline might see that a similar discussion already took place, and would therefore suggest that the participants look at what was already said.

Lloyd wrote:Something I like in your theory is showing that there is only one alternative that's workable. [...] It seems to me that that is the Occam's razor approach, which has the advantage of simplicity, cutting off all the stubble theories.

Yes, and this is what we get when we lay out all of the theories, and we get to compare their features. Every single one of them can explain something. But celestial objects are not made one way so that one aspect can be explained with one theory, and then reforming themselves out of new stuff so that some other aspect can be explained with a different theory -- all of the properties of a celestial object are (somehow) being generated by the same stuff. ;) So each property can be used as an individual test of a theory, while the successful candidate has to pass all of the tests. ;)

Lloyd wrote:Don't you all think it might be helpful to have a logic tree for all of science?

Yes.
Give a man a fish and you feed him for a day. Teach a man to fish and he'll spend the rest of the day sitting in a small boat, drinking beer and telling dirty jokes.

Volcanoes
Astrophysics wants its physics back.
The Electromagnetic Nature of Tornadic Supercell Thunderstorms
User avatar
CharlesChandler
 
Posts: 1790
Joined: Tue Aug 19, 2008 6:25 am
Location: Baltimore, MD, USA

Re: Online scientific discourse is broken and it can be fixe

Unread postby Lloyd » Sun Dec 08, 2013 12:58 pm

CC: I prefer my site, because it's more flexible, and I don't think that we're sure what we want yet. ;)

I'm sure of what I'd like to try out, because it looks the most promising. I'd like to start with truthmapping, but with the features that you'd like included added. Here's where I started a logic map for Science, although I called it for Astrophysics, I think: http://truthmapping.com/viewtopic.php?id=1328.
LK: But I'd like to see a feature like on truthmapping for critiques and rebuttals.
CC: if it has an advantage, let's see if we can instantiate that in QDL. Is it that you want the premises, critiques, and rebuttals clearly labeled (like issues, ideas, pros, & cons in Deliberatorium)? I was thinking of making the icons in QDL user-selectable. So you could go with the standard gear icon, or you could select a different icon, to designate what kind of item it is. (BTW, who trashed all of the stuff that I entered into the Deliberatorium?)

I didn't know you posted anything on Deliberatorium. I only saw the rebuttal you posted to a critique on my Formation of the Sun map on truthmapping. I didn't see your rebuttal last time I was there, so I don't know if it was deleted, or if it's hidden until a button gets clicked.

What I like at http://truthmapping.com/viewtopic.php?id=1328 is:
1. the Map on the left;
2. the list of statements on the right
3. the dialog boxes for each statement on the right for Critiques and for Rebuttals
4. and the Rating option for each Critique, Rebuttal and Statement.

Can you set up a page on your site with those features? And then add the extra features you mentioned?
LK: Comparing a bunch of disparate theories seems too overwhelming.

As long as there's the option to hide everything one doesn't want to look at, like the way you have your site set up, it should be okay.
LK: Don't you all think it might be helpful to have a logic tree for all of science?
CC: Yes.

Does the logic tree I started in the link above look like an okay beginning? It doesn't have much branching yet, but I expect it will, if we work with that or something similar.
Lloyd
 
Posts: 4317
Joined: Fri Apr 04, 2008 2:54 pm

Re: Online scientific discourse is broken and it can be fixe

Unread postby pln2bz » Sun Dec 08, 2013 2:39 pm

Truthmapping seems to be a step in the right direction, insofar as it switches the dimension for user contributions from one that is time-based to one that is topic-based. I completely agree that this is a necessary step for scientific discourse. Think about how much time is wasted every day by people trying to make a complex scientific case, just because they are being forced to start from scratch each time they try to make the same case. When we think about science, we do not start from scratch each time we encounter a familiar topic. Our thoughts build from and upon our pre-existing knowledge of that topic. With this conventional way of discourse, the more divergent a new idea is, the more explaining would be required to justify it. So, I would argue that the conventional thread-based communication tools are generally failing to represent/service new ideas in science, and that this necessarily has an impact upon society's perception of new ideas in science. Oftentimes, emergent scientific ideas promoted on forums (etc) will be ignored, in part because those who are promoting them are forced to start at the beginning each time, even though they do not even know what the "beginning" actually is for their readers, and even though it probably differs for each of their readers.

So, notice what is happening here: There is a group of people -- people who think and talk about emergent ideas in science -- who have a specific need: We need a system which facilitates the discussion of new, complex scientific ideas. In the worlds of marketing and user interaction, this topic-based discussion can be called a "need state". The Truthmapping video offers an excellent example for how the EU community should be thinking about this problem, insofar as they explain how their tool addresses four distinct need states: the need to avoid threads unrelated to topic, to visually de-value "aha moment" discourse, to resolve the Soapbox Problem and to expose hidden assumptions.

What I would like to suggest, however, is that the actual total set of problems which we can identify which need to be resolved by a scientific social network -- based upon our experiences as Electric Universe advocates, and research into science education (etc) -- is actually quite a bit larger than these four needs. I am already up to 28 distinct need states, and I expect to discover many more as I learn more (I will post these shortly, but I also want to encourage people to try to think about this on their own first …)

I suspect that it is at this step of compiling the need states that many people start to feel overwhelmed about a large project like this. And so, what I would like to encourage (for people who are interested in this idea of creating a scientific social network) to do is to get organized. Create your own need state document and add to it each time that you think of a new need state (If you happen to have a new iPhone with iOS7, you can for instance install Evernote for free, which will then let you add to your list wherever you go, and even dictate it by voice if you like ...). The site design should be a direct response to the needs of actual site users.

In an actual business model, there would be an extra step layered on top of that: The site users would be broken down into categories of users called "user personas" … PhD students, high school students, entrepreneurs, teachers, etc … and the need states would be defined independently for each of those user categories. And then, to create an effective business plan, a strategy would be formulated for addressing the most profitable user category to focus upon. That is not to say that other user categories would not be addressed; what it means is that an actual business would try to absolutely nail a single user category, while also peripherally addressing a number of other user categories.

Once you start thinking in terms of need states and these user personas, and you review everything that is out there, what I think that most people who think about these things will notice is that none of the options out there completely satisfy all of the needs which we are aware of (yet). But, there is certainly value to listing out the need states which various sites appear to satisfy, and comparing and contrasting our own lists.
pln2bz
 
Posts: 248
Joined: Sun Mar 16, 2008 8:20 pm

Re: Online scientific discourse is broken and it can be fixe

Unread postby CharlesChandler » Sun Dec 08, 2013 11:33 pm

@ pln2bz:

I totally agree with what you're saying here! The TruthMapping video does a great job of identifying the problems of journal-style discussions. I just think that their site is hard to read, and there are far too few features. Also, as a software developer with over 25 years in the business, I can tell you that software with an extremely rigid workflow (like TruthMapping) frequently makes a nice demo, but rarely survives in the real world. Unless you have a very specific task that you do quite frequently, and that requires specialized software, you don't need a purpose-built app. Rather, you need something that is flexible enough to bend to different types of problem-solving situations. To think that we could hammer the scientific discovery process into a rigid workflow is actually a bit far-fetched. Lloyd & I have been working for a couple of years now on trying to define a process, and no matter what we do, the process that we define works for the problem we had in mind, but not for the next problem!!! :)

Recently we've been reviewing debating software (TruthMapping, Deliberatorium, TopicCentral, and DebateGraph). All of them are addressing the same problem: great ideas are getting buried in online discussions, and the ideas need to be laid out in a logical way, not a chronological way. Each site has cool features, but they all have very limited capabilities, and very rigid workflows. So I've been making improvements to my site, which I think has more potential. The most important feature is just the ability to express the concepts in an outline format, which my site does (along with a lot of other stuff). So here's what this looks like on my site. The icons identify the type of item (the gear icon is just a generic marker; the red exclamation point is a problem; the lightbulb is an idea, and the key is a solution). There are over 30 icons to choose from, for whatever type of stuff you're doing, so those aren't the only ones -- those are just the ones that came into play on this particular topic. The green arrows are "disclosure widgets" that allow you to expand/collapse items, to get more detail, or to go back to overview mode. These don't do anything in the following link, which is just a save-as from my site -- you have to register and log in for these to work.

Outline

I'd like to suggest that you register, and then you can enter your "28 distinct need states", and we can collaborate on defining the requirements for community software. You should be able to use topic-based software to lay out the requirements for decent topic-based software, right? :) Anyway, my site supports a lot of features, including threads, articles, surveys, outlining, bibliographies, etc. It also has a rights system, so that people can define who gets to see or even edit their material. So it's a full featured package, and online collaboration is what it was designed to do, so please check it out. (Registration is free.)

http://qdl.scs-inc.us/
Give a man a fish and you feed him for a day. Teach a man to fish and he'll spend the rest of the day sitting in a small boat, drinking beer and telling dirty jokes.

Volcanoes
Astrophysics wants its physics back.
The Electromagnetic Nature of Tornadic Supercell Thunderstorms
User avatar
CharlesChandler
 
Posts: 1790
Joined: Tue Aug 19, 2008 6:25 am
Location: Baltimore, MD, USA

Re: Online scientific discourse is broken and it can be fixe

Unread postby pln2bz » Mon Dec 09, 2013 10:05 am

CharlesChandler wrote:I just think that their site is hard to read, and there are far too few features. Also, as a software developer with over 25 years in the business, I can tell you that software with an extremely rigid workflow (like TruthMapping) frequently makes a nice demo, but rarely survives in the real world. Unless you have a very specific task that you do quite frequently, and that requires specialized software, you don't need a purpose-built app. Rather, you need something that is flexible enough to bend to different types of problem-solving situations. To think that we could hammer the scientific discovery process into a rigid workflow is actually a bit far-fetched.


Hi, Charles! It sounds like you guys have already familiarized yourself with the debate mapping graveyard. This has been going on for decades now, since the very beginning of the Internet. For example, take a look at what was probably the first computer-based attempt to solve this problem: Robert E Horn's Argumentation Mapping Project. Those attempts are very educational, for the way in which they tend to fail reveals an important detail about how to move forward.

Do you notice the whitespace problem? Our vision system uses whitespace to delineate categories. This is a fundamental graphic design principle. Whitespace is not the enemy; it's the canvas which the mind uses to delineate things. The mind has to be trained to read text, but it can identify objects without any observable effort at all. So, as you have apparently noticed, it seems that what we need to be doing is attempting to emulate the way in which the mind actually represents and identifies concepts.

Peter Gardenfors has taken this idea of space between objects a step yet even further. His book Conceptual Spaces makes a case that the way that our mind encodes abstract concepts is actually just a manifestation of the way we do navigation. This is an appealing idea (which could nevertheless turn out to be wrong, of course) in terms of cognitive evolution. It's a very interesting proposition because much progress has actually already been made in understanding spatial cognition (in rats' brains, at least). The point is that, if Gardenfors is right, then conceptual typicality (for example, a bean bag is less of a chair than an office chair ...) could actually turn out to be function of the actual geometry involved with the encoding of the concepts in the mind. It would also suggest that we might learn some things about conceptual encoding by studying the empirical results that relate to spatial encoding.

Now, take a look at Rob Spencer's fractal map visualization (make sure you try clicking the maps, and changing the visualization style to blocks). This is not meant to suggest a specific solution, but rather a general approach of using fractals and depth (much like Prezi) as a dimension for visualizing relations between topics. And notice that he was able to do this with HTML5's canvas (!). In the event that a large site was built, this would be a very important detail, because Google's search engine would be capable of crawling the content. Building the site is just one aspect of the problem; another critical detail is making sure that the search engine optimization (aka SEO) is done properly so that the topics score well in Google's rankings.

Another topic that I'll be getting to pertains to the way in which these topics, and relations between topics, are encoded into the database. Rather than using a traditional relational database, it becomes apparent with a little bit of research that a graph database is far more suited to working with scientific structure. Relational databases like SQL run into very serious query latency problems with lots of SQL table joins. We want something which will scale as the number of users rises, so the graph database is the way to go.

Charles, you might also want to tune into the work of David Blei and this notion of topic modeling. I'll be attempting to explain the mechanics of it as time permits, but if you are the point where you are engaging the visualization aspects of this project, it's something you might want to invest some time into understanding, because topic modeling permits us to take a document or user contribution and automatically extract the topics on the basis of context. This is important because it permits the computer to understand the difference between plasma (the fourth state of matter) and plasma (in the blood). Topic modeling is a very challenging subject, since it is basically machine learning applied to documents. From what I can tell, machine learning is basically just a system for optimizing algorithms using statistical tools. What you will find, if you choose to look into it, is that you don't actually have to understand this down to the lowest levels so long as the libraries which are out there satisfy your needs. Most of these more technical libraries will be written in Python.

Visualization is really quite critical, and I would advise that people keep an open mind to new ideas about how to pull it off. I do appreciate action-oriented programmers who want to get their hands into coding as quickly as possible, but I'm also cognizant of all of the needs associated with an adaptive solution. In order for the system to adapt to a wide variety of problems, we need to develop familiarity with all of those problem types, and all of the technologies which are currently trending.
pln2bz
 
Posts: 248
Joined: Sun Mar 16, 2008 8:20 pm

Re: Online scientific discourse is broken and it can be fixe

Unread postby Lloyd » Mon Dec 09, 2013 10:07 am

Seeking Efficiency
I guess we're all trying to do that.

What do yous think of this? The most efficient way to evaluate scientific papers and theories is to find the vital statements and determine if any of the vital statements are disproven. If none are disproven, then the paper or theory is possible. Otherwise, it is archived.

It's pretty easy to break down a paper or theory into individual statements, using a Word program or the like. The next step would be to look for the vital statements of the theory and label them as Known, Disproven, or Unproven.

If we construct a library of scientific knowledge, all theories that are thus disproven go to Archives.

Theories that are Unproven go to the Unproven section of the library.

And the theories whose vital statements are all Known to be True go to Scientific Knowledge.

Independent vs Collaborative Evaluation
Most of us do Independent evaluation of theories along with inefficient Collaborative evaluation. The forum here doesn't have a way to do effective collaborative evaluation of theories. Everyone has their own ideas, but no collective evaluations. It seems that even the EU team has few collective agreements.

My suggested process above would require collaborative evaluation, so our ultimate Science Program should seek to make that efficient, I think.

Each collaborator could evaluate different papers/theories, finding the vital statements and labeling them as Known, Disproven, or Unproven. Then other collaborators could evaluate each of those evaluations and say Agree or Disagree.

Real Time Debates
Debates can improve efficiency too. The programs we've been discussing here lately are geared in that direction, I think. And TruthMapping seems to have the best features for that so far. But Debating may not need to be something to focus on yet. Maybe finding efficient means to evaluate theories would be more productive. When there are disagreements about evaluations of vital statements, TruthMapping Debates might be more productive then.

Charles has listed a bunch of theories at QDL, so that might be a place to start evaluation. And anyone would be able to post other theories there too which anyone could then evaluate.
Lloyd
 
Posts: 4317
Joined: Fri Apr 04, 2008 2:54 pm

Re: Online scientific discourse is broken and it can be fixe

Unread postby Plasmatic » Mon Dec 09, 2013 3:07 pm

Chris said:

Do you notice the whitespace problem? Our vision system uses whitespace to delineate categories. This is a fundamental graphic design principle. Whitespace is not the enemy; it's the canvas which the mind uses to delineate things. The mind has to be trained to read text, but it can identify objects without any observable effort at all. So, as you have apparently noticed, it seems that what we need to be doing is attempting to emulate the way in which the mind actually represents and identifies concepts........
Peter Gardenfors has taken this idea of space between objects a step yet even further. His book Conceptual Spaces makes a case that the way that our mind encodes abstract concepts is actually just a manifestation of the way we do navigation. This is an appealing idea (which could nevertheless turn out to be wrong, of course) in terms of cognitive evolution. It's a very interesting proposition because much progress has actually already been made in understanding spatial cognition (in rats' brains, at least). The point is that, if Gardenfors is right, then conceptual typicality (for example, a bean bag is less of a chair than an office chair ...) could actually turn out to be function of the actual geometry involved with the encoding of the concepts in the mind. It would also suggest that we might learn some things about conceptual encoding by studying the empirical results that relate to spatial encoding


I think that this is a good direction but have some distinctions to make. We use words as epistemological equivalents to concrete entities. Concepts are integrations of sense data and involve abstracting from particulars/entities to form units or categories. The mind cannot hold this open ended grouping in one frame so we economize this integration with a symbol. The problem is the structure of predications and logical hierarchy of concepts and premises is not easily determined via written format. I could lay out in detail the epistemology behind this process.(without an EEG...)

This is why I prefer concept maps. The logical priority-relationships is visually represented. One has no difficulty determining the top of a pyramid or the tip of a tree. A pictorial map which says you are here with an arrow and an x, is much easier to comprehend, than say, a written description like," you are located in the fourth row of the second story pavilion area... But language enables us to let go of perceptual images by economizing those referents into a single concrete word. This is a requirement of the limitation of the bandwidth of consciousness.

I wanted to upload one of my concept maps but the attachment feature won't accept my Iphone image.
"Logic is the art of non-contradictory identification"......" I am therefore Ill think"
Ayn Rand
"It is the mark of an educated mind to be able to entertain a thought without accepting it."
Aristotle
Plasmatic
 
Posts: 800
Joined: Thu Mar 13, 2008 11:14 pm

Re: Online scientific discourse is broken and it can be fixe

Unread postby pln2bz » Mon Dec 09, 2013 3:22 pm

Lloyd wrote:It's pretty easy to break down a paper or theory into individual statements, using a Word program or the like. The next step would be to look for the vital statements of the theory and label them as Known, Disproven, or Unproven.

If we construct a library of scientific knowledge, all theories that are thus disproven go to Archives.

Theories that are Unproven go to the Unproven section of the library.

And the theories whose vital statements are all Known to be True go to Scientific Knowledge.

Independent vs Collaborative Evaluation
Most of us do Independent evaluation of theories along with inefficient Collaborative evaluation. The forum here doesn't have a way to do effective collaborative evaluation of theories. Everyone has their own ideas, but no collective evaluations. It seems that even the EU team has few collective agreements.

My suggested process above would require collaborative evaluation, so our ultimate Science Program should seek to make that efficient, I think.

Each collaborator could evaluate different papers/theories, finding the vital statements and labeling them as Known, Disproven, or Unproven. Then other collaborators could evaluate each of those evaluations and say Agree or Disagree.


You seem to be focusing upon the easy stuff first. I would recommend focusing upon the harder gray areas. How can we make progress towards consensus in the process of evaluating models when the models oftentimes originate from competing worldviews? I believe that this problem suggests a need to distinguish the model and worldview levels of discourse. And if you pay close attention to sites like Slashdot and physorg, I think what you will generally observe is the presence of model-level thinkers who prefer not to engage in worldview-level discourse (and I think we can now see that there are good reasons why they are averse to it).

So, part of the difficulty here is in decoding all of these things which are all happening at the same time on the same communication channel. In electrical engineering, this is what we would call performing a fast-fourier or Laplace transform: It's a process of pulling all of the component signals out of the original signal. Unfortunately, there is no system of equations to help us here in this instance. We're going to have to do this manually, and it will be a very messy and difficult process which will require lots of observation, interaction and hopefully, one day, actual testing. Some people call these "wicked problems", by the way; they are problems which have no known solutions, and which there are no simple processes for solving.

The System of Systems Concept of a Scientific Social Network

Here's a clip from my own notes which strives to explain the system-of-systems idea for a scientific social network, which I think more accurately reflects how our social institutions actually handle complex problems. Remember: What we want to do here is to emulate that which works in the real world. But, where we see something which could use an upgrade, we want "levers" which we can use to shift the discourse into a particular direction. I use the term "values" here as an intentionally ambiguous construct which might be enacted through some system of user ratings. But, as I've hinted at previously, we'd be wise to treat the system of user ratings separately and with care, for we have options there too ...

Since science exhibits need states which, if permitted, can clash with one another, scientific dialogue should be viewed as a system-of-systems rather than one single coherent two-way broadcast system. The need for accuracy in claims, for instance, is plainly in conflict with the need for creative suggestions. If the technical number cruncher is permitted, he will take a shot at some whimsical suggestion made by the artsy guy. But, this is not to say that neither creativity nor the arts have no place in science -- especially in terms of how we convey and teach it.

The reason why I would not personally spend a whole lot of time trying to perfect a mailing list is that the competing need states will tend to obscure one another implicitly through the dialogue. One person might scold another, for instance, for some transgression -- and suddenly an implicit reactionary feeling spreads through the entire list that the scolded behavior was something which others will now actively avoid. The problem is that the reaction is almost always an overreaction which might only be temporal, but which generally creates inconsistency in the values that guide the dialog. However, from what I can tell, the values which are associated with thinking like a scientist should generally remain fixed and directed, and the way to do that is to create places where they permanently reside, without having to directly fight one another for survival.

The words we use derive much of their meaning from context. When two people disagree about a claim, the disagreement can be related to the claim itself; but oftentimes, the unique knowledge and experience which each person brings to the dialogue can also contribute to any disagreement. In discussions of science, I would propose that there exists four separate modes of dialogue which derive from the structure of science itself:

  • There are discussions of concepts, since these meanings can vary over scales, domains of inquiry and based upon competing worldviews. For instance, the concept of wetness loses its meaning once the scale is only slightly shifted from that of water to its constituent, hydrogen. The concept level should perhaps be most friendly to the layperson's needs. Those who are willing to help others by mentoring them might, for instance, be most valued at this level.
  • Relationships between concepts are called propositions. When a proposition applies to more than a single instance -- when it is generalizable -- we call it a principle. Propositions exhibit a continuum of certainty, from the least certain (focus questions) to most certain (claims & principles). The proposition level could, as an example, be designed to be a place where it's okay to ask questions. It's also where creativity (divergence) might be most valued.
  • A model or "theory is a set of interrelated principles that enables explanations or predictions of interactions among objects & events" (The Art of Educating with V Diagrams, page 53). Thus, we can also have discussions about competing models within just one worldview, and we can choose to discuss these models in terms of either their conceptual propositions, or more precisely, their mathematics. Many science journalists, professors and textbooks today unfortunately cast these discussions as though they are the highest-level discourse we can have in science. The model level is meant to be a place where theories can be cultured by theorists, experimenters and promoters who are already largely fluent in the model. Therefore, it might be considered to be a place where consensus (convergence) is most valued and credentials can matter, but those same credentials might be socially meaningless from the perspective of a different worldview.
  • And yet, look at the comments attached to scientific press releases today on popular science aggregator websites, and it's easy to see that discourse can also occur at a level beyond the models themselves. Models tend to originate within the belief systems -- the worldviews -- of scientists, and these worldviews are oftentimes taught by universities as part of the curriculum. These tend to be the most difficult conversations to have, for they require a deep familiarity with multiple competing worldviews. That said, there can be incredible value to discussing competing models which stem from differing worldviews, because even if it might be widely believed that a particular model is the best, there is still the possibility that a better model might originate from a competing worldview. In philosophy of science, this is called the problem of unconceived alternatives. It offers us a reason to believe that all scientific theories are provisional, and it is not necessarily a solvable problem. We are possibly forever stuck with the possibility of unconceived alternatives. The worldview level is a place where people of differing worldviews can critique each others' competing models. This is perhaps where most controversy should be situated. Those who demonstrate either a wide breadth or deep specialty in knowledge are most valued. Users are encouraged to detail their knowledge base coverage. But, notice that nobody can expect to know it all, and so this level must value both cooperation and competition (which some call "co-opetition")


Here's a helpful quote which suggests that we have competing forces which all need to be taken into account in whatever system we design ...

"An Educational Model

'the sensible educator. . . will not expect or intend to produce an educated adult who has no beliefs, values, or attitudes, which he cannot rationally defend against all corners and who is incapable of settled convictions, deep-seated virtues, or profound loyalties. But neither will he treat his pupils in such a way as to leave them with closed minds and restricted sympathies. The process of being educated is like learning to build a house by actually building one and then having to live in the house one has built. It is a process in which the individual inevitably requires help. The extreme authoritarian helps by building the house himself according to what he believes to be the best plan and making the novice live in it. He designs it in such a way as to make it as difficult as possible for the novice to alter it. The extreme liberal leaves the novice to find his own materials and devise his own plan, for fear of exercising improper influence. The most he will do is to provide strictly technical information if asked. The sensible educator helps the novice to build the best house he can (in the light of accumulated experience). He strikes a balance between the need to produce a good house and the desirability of letting the novice make his own choices; but he is careful that the house is designed in such a way that it can subsequently be altered and improved as the owner, no longer a novice, sees fit.'


Professor Basil Mitchell
The Durham Report
(excerpted from Beliefs and Values in Science Education" by Michael Poole)


Although dialogue within differing levels should be expected to refer to other levels, a particular dialogue should perhaps be fundamentally rooted within just one, based upon a person's needs -- be it learning, asking, creating or debating. By creating a system of weakly interacting systems (aka "channels" or "layers"), each with its own set of (possibly competing) values, no single such purposeful activity would ever be permitted to obstruct any of the other purposeful activities. The mind automatically recognizes that since each of these sets of values derives from an established channel, that those values should always be present. We simply choose which to pay attention to at a particular moment, based upon the context of our cognition.

Thus, I've identified at least four -- and there are possibly more (?) -- fundamental levels of discourse which we can have in science. In practice, mixing these levels of discourse can cause confusion or frustration. Trained scientists might, for instance, prefer to take for granted some set of assumptions in order to focus, at a particular moment, upon the creation of more predictive models based upon a particular worldview. At that particular moment, having somebody in the room who insists upon questioning the worldview itself could be an annoyance (and understandably so …).

A common problem arises in the reverse as well, however, when scientists refuse to mention to the public anything about the assumptions or hypotheses which might be associated with the worldviews upon which their models are based. At the point where such premises are widely taken for granted, then even though they lack scientific basis, they can nevertheless become culturally enshrined as facts -- which is counter to traditional scientific norms. Thus, there exists a cost associated with the rigor of our scientific methodology, as well as progress in science, when it becomes fashionable to pretend as though there exists no clash of worldviews in science worth discussing. The fact is that there exist many debates in science that have lasted for many decades outside of the textbooks that are used to train university students. In some cases, the Internet is the only place where such discussions continue on.

As an example, Tom Van Flandern points out that -- at least, when he checked -- the most frequently asked question on the topic of gravity in USENET groups properly reflects a question which has been asked in print "thousands of times in dozens of different ways over the years since general relativity (GR) was set forth in 1916". That question would be "What is the speed of gravity?" (See "The Speed of Gravity: What the Experiments Say"). Nevertheless, it would seem that this seemingly important question is apparently not generally explored within our science classrooms or textbooks. The question definitively demonstrates the importance of considering worldviews in science -- a practice which textbooks have generally not been designed to service -- since the answer is generally determined according to the constraints of the worldview itself.


What is nice is that (a) this is exactly the type of problem that graph databases were designed to solve, and (b) we can treat the manner in which these different levels of discourse "weakly" interact as a visualization problem. And for our purposes, that means that we need not necessarily think too much about it just yet.
pln2bz
 
Posts: 248
Joined: Sun Mar 16, 2008 8:20 pm

Re: Online scientific discourse is broken and it can be fixe

Unread postby Plasmatic » Mon Dec 09, 2013 4:14 pm

Just wanted to note that the political aims of constructivism is particularly visible in those quotes.
"Logic is the art of non-contradictory identification"......" I am therefore Ill think"
Ayn Rand
"It is the mark of an educated mind to be able to entertain a thought without accepting it."
Aristotle
Plasmatic
 
Posts: 800
Joined: Thu Mar 13, 2008 11:14 pm

Re: Online scientific discourse is broken and it can be fixe

Unread postby CharlesChandler » Mon Dec 09, 2013 7:37 pm

pln2bz wrote:@Lloyd: You seem to be focusing upon the easy stuff first. I would recommend focusing upon the harder gray areas.

Well, OK. But somewhere in here, we need to take a step back, and realize what we're doing.

We have all acknowledged that online discussions, in their journal-style chronological accumulation of thoughts, have severe problems. Decent ideas get buried, because the conversations wander off in new directions, and except to those directly involved in the conversations, the value of the decent ideas becomes inaccessible.

We have discussed the problems inherent in discussions. And we have discussed them some more. And we keep discussing them. And all in all, it seems that we're just spinning our wheels here. Why? Because online discussions have severe problems -- they don't tend to lead anywhere, except round and round. :D And if we're going to keep debating the issues at the highest of philosophical levels, I guarantee you that this will never lead anywhere!

So I'm building an outline of this discussion, using my software. :D I'd like to suggest that people participating in this thread familiarize themselves with it.

http://qdl.scs-inc.us/?top=5960

If you register an account, you get the advanced navigation tools, and you can add stuff yourself, so if you're interested in contributing to this, I'd highly recommend registering. (It's free.)

So, we should be able to use debating software to debate the requirements of debating software, right? :) Lloyd and I have been doing this for a couple of years now, and many of the features got implemented to support our debate on how to debate stuff. ;) But I'm not suggesting that this has to be used in a fancy way. Rather, I'm just using it for its ability to do expandable outlines. In outline format, everything gets sorted out into proper places, fulfilling the design requirements of eliminating flooding and off-topic meandering, revealing unstated assumptions, and suggesting interpolations/extrapolations of what's already there. The ability to expand and collapse items makes it easy to drill down for more detail, or to go back up to the top level and skim over the entire thing. I really like the way the structure of these ideas is emerging, so I'm going to continue with it. Flat lists are OK for relatively small issues, but anything of any complexity at all requires a way of sorting everything out, and an expandable outline provides that.

Philosophically speaking, this might not be the final solution, and time permitting, I'm willing to consider other things. I'm familiar with the type of code that DebateGraph uses, since at my previous job, I did a bubble diagram module for architectural design, where the bubbles represent proposed rooms in a building, and the architect shuffles them around to get related rooms clustered together before beginning work on the actual floorplan. I also did a Visio-style block diagram module, for logical flowcharting.

Over the years, I have attempted to use such technologies in the various intellectual endeavors that consume my free time, but I have never stuck with any of them. I think that my biggest problem with block diagrams is just that there really isn't very much information in view at any one point in time, and the extremely narrow perspective detracts way too much value. Though I think visually, and my professional expertise is in graphics programming, I always lay out ideas in text, and if there is any significant structure to it, I'll do it in outline format. It's just faster that way, and I think that most people would agree. Block diagrams are time-consuming to produce, and only convey a few simple ideas. They're extremely useful for presentations. But nobody seems to use them for capturing and nurturing the initial thoughts, despite how much effort has been put into developing such technologies. My opinion is that if this was the way to go, we'd already be using Visio for everything (on our desktops, or online). There's a reason for why we aren't. Why is that? ;)

Anyway, I'd like to encourage others to map this discussion in other software, so that we can all evaluate what it looks like. I'll do it in my software, just as an expandable outline. People who want to champion the use of different packages might just use the work that I've done in sorting it out as their starting point.

But I find it to be a little senseless for us to continue this discussion of the futility of time-based statements using time-based software! I believe that forums are great for capturing ideas and getting quick feedback. But if the ideas are not subsequently plugged into a topic-based system, they're gone forever. Thus we should keep discussing the issues, but we should also start building and evaluating the structure that is emerging from these great ideas. ;)
Give a man a fish and you feed him for a day. Teach a man to fish and he'll spend the rest of the day sitting in a small boat, drinking beer and telling dirty jokes.

Volcanoes
Astrophysics wants its physics back.
The Electromagnetic Nature of Tornadic Supercell Thunderstorms
User avatar
CharlesChandler
 
Posts: 1790
Joined: Tue Aug 19, 2008 6:25 am
Location: Baltimore, MD, USA

PreviousNext

Return to The Future of Science

Who is online

Users browsing this forum: No registered users and 2 guests