Cognitive functions

[AtheisticNonduality]

2 minutes ago, Carl-Richard said:

You sure? if I were to take your previous post as something more than just virtue signalling, you seem to not understand the basic distinction between criticizing cognitive functions and typology (I'm only doing the latter).

Cognitive functions are like parts to the whole of typology.

[Carl-Richard]

1 hour ago, AtheisticNonduality said:

Cognitive functions are like parts to the whole of typology.

There are in-built claims about correlations between behaviors that need to be justified somehow:

For example, Ti users will also use Fe, like Fi users will use Te. Ne users will also use Si, like Ni users will use Se. Frequency of one determines the frequency of another.

This goes beyond pure descriptions and into predictions. It's fine to say "this guy is thinking about zebras right now", but to then say "that means he will think about unicorns later" is a step up from that.

[thisintegrated]

15 minutes ago, Carl-Richard said:

There are in-built claims about correlations between behaviors that need to be justified somehow:

For example, Ti users will also use Fe, like Fi users will use Te. Ne users will also use Si, like Ni users will use Se.

This goes beyond pure descriptions and into predictions. It's fine to say "this guy is thinking about zebras right now", but to then say "that means he will think about unicorns later" is a step up from that.

At least you're actually debating now.

imo, the 4 letter system needs a complete overhaul.  Maybe a move to a 5/6 letter system would be good.  P and J are unnecessarily confusing, especially to new people.

And the function should, imo, be used in order of strength.  Saying "Ne means no conscious Te" is too ambitious.  Only a supercomputer could come up with such strict rules/connections.  I've seen no real evidence that "Ne = no conscious Te".  I should be able to say I'm a NeTiNiFe, for example.

ime, strong Ne = strong Ni.  But this fact is not addressed.  Anywhere.  Ever.  So the most advanced MBTI users end up combining both, the 4 letter system and the 16 letter system.  The fact that someone is an "xNxx", by itself, is significant.  Even if the functions say "good Ni = bad Ne", a strong N is a strong N.

 

Now, although it's bad and confusing, it still works.  And there exists no better alternative.

Edited April 18, 2022 by thisintegrated

[Carl-Richard]

@thisintegrated I appreciate the openmindedness #Big5lingo ?

I might be critical of it, but I truly wish there existed a perfect typology.

[thisintegrated]

3 minutes ago, Carl-Richard said:

@thisintegrated I appreciate the openmindedness #Big5lingo ?

If this was an MBTI forum, I'd be super critical of MBTI.

I adapt to where I am??

 

No one knows my actual stance on anything?‍♂️

Edited April 18, 2022 by thisintegrated

[no_name]

22 hours ago, Carl-Richard said:

That's also my view, but this changes when you start talking about typology (and claims about the frequency one uses the functions). Once you start talking about correlations to behavior (and correlating the functions with each other), you're in the realm of science. What are the justifications for combining these functions into types?

Logically I would think the justification is as follows: According to Carl Jung, the 4 functions are Sensing, Intuition, Feeling and Thinking. Each person must have these 4 functions - you can’t have a person who doesn’t use thinking or feeling. You can’t use these function in equal amounts 25%-25%-25%-25% throughout your day, if I was to follow you for n days and record these percentages, I would get a breakdown of what your natural preferences are.

If you are indeed studying psychology and expect it to be an exact science, you are in a wrong field and should switch to actual hard sciences instead. There is a lot of judgement and not easily quantifiable concepts, many are not even well defined and are still in the process of defining. 

While big 5 have a basic statistical model, all what it’s doing is finding correlation between words, big 5 concepts are nothing useful, they are just words. What use will it do for you to know that you are conscious or neurotic? What do these things even mean? 

My original post mentioned that I don’t know why the concepts of cognitive functions are not studied more, why are there no studies about this, why are there no proper tests created to identify those functions? Is it because it’s too complicated? There is no reason why a cluster analysis, regression analysis or some decision tree can’t be built around these concepts.

[no_name]

8 hours ago, thisintegrated said:

I've seen no real evidence that "Ne = no conscious Te".  I should be able to say I'm a NeTiNiFe, for example.

So you don’t believe the +-+- system has any grounding to it? I don’t see how someone could be Ni and Ne or Ti and Te, in my head there exists a clear preference.

Also in your example there someone who has Ne and Ni as their first 3 functions, wouldn’t that person be some kind of hallucinating psycho? Don’t they need an S function to survive? 

As an Ni dominant I appreciate and understand Ne, but it gives me anxiety if I had to do it and I don’t like doing it. E.g., if you were to ask me to brainstorm, I would get tired of it and start clinging to some one item instead that I find most interesting, yet I do like it when someone does the brainstorming for me and shows me other options 

I do not understand why you can’t have an NiFiTiSe for example though, but I am still just learning so maybe there is a reason..

Edited April 18, 2022 by no_name

[Carl-Richard]

 

4 hours ago, no_name said:

Logically I would think the justification is as follows: According to Carl Jung, the 4 functions are Sensing, Intuition, Feeling and Thinking. Each person must have these 4 functions - you can’t have a person who doesn’t use thinking or feeling. You can’t use these function in equal amounts 25%-25%-25%-25% throughout your day, if I was to follow you for n days and record these percentages, I would get a breakdown of what your natural preferences are.

I know you're going to address this at the bottom of your post, but I'll just repeat it:

These are intuitions about correlations of behavior. If they make intuitive sense, that means they should be tested to see if the intuitions don't contradict the empirical "reality".

 

4 hours ago, no_name said:

If you are indeed studying psychology and expect it to be an exact science, you are in a wrong field and should switch to actual hard sciences instead. There is a lot of judgement and not easily quantifiable concepts, many are not even well defined and are still in the process of defining. 

Yes, generally speaking, psychology is a fluffy science, but we still choose the least fluffy models when we can.

 

4 hours ago, no_name said:

While big 5 have a basic statistical model, all what it’s doing is finding correlation between words, big 5 concepts are nothing useful, they are just words. What use will it do for you to know that you are conscious or neurotic? What do these things even mean.

Just with cognitive functions, you have to learn what the traits mean (descriptions of behaviors and correlations between behaviors). Each 5 traits contain 6 sub-traits within them (for a total of 30). This means you can get 30 scores to describe your personality. Somebody who gets a similar score to you would have to answer the same things on the test, and those questions are supposed to represent the constructs in question (traits/sub-traits). It's crude and limited, but according to the experts in personality psychology (yes, this matters), it's the best thing we've got.

 

4 hours ago, no_name said:

My original post mentioned that I don’t know why the concepts of cognitive functions are not studied more, why are there no studies about this, why are there no proper tests created to identify those functions? Is it because it’s too complicated? There is no reason why a cluster analysis, regression analysis or some decision tree can’t be built around these concepts.

?

I'm not saying absense of evidence is necessarily evidence of absense, but maybe it is in this case? ?

[no_name]

1 hour ago, Carl-Richard said:

I know you're going to address this at the bottom of your post, but I'll just repeat it:

These are intuitions about correlations of behavior. If they make intuitive sense, that means they should be tested to see if the intuitions don't contradict the empirical "reality".

I will address it on top.

You are misusing terms again. Correlation measures relation between two variables. I am talking about proportions. 

1 hour ago, Carl-Richard said:

Yes, generally speaking, psychology is a fluffy science, but we still choose the least fluffy models when we can.

1 hour ago, Carl-Richard said:

I'm not saying absense of evidence is necessarily evidence of absense, but maybe it is in this case? ?

I think the problem is that there are not enough statisticians and data scientists involved in the field of psychology. Running these kind of experiments is a piece of cake for a data scientists, but psychology researches lack knowledge in these areas, so it requires more effort. Data scientists would also need the input of psychologists to understand the terminologies and to be able to design the experiment properly. There should be more inter-weaving of the two fields.

Also, no offence, but talking to you feels a bit like this right now:

You: This wouldn’t work because the grass is purple, the experiments show that the grass is purple.

Me: The grass is green, it’s just that when you add certain chemical to the fertilizer the grass turns purple. 

You: Yea, the grass is purple and it’s the best what we’ve got so far, I see all this purple grass so I can verify that the grass is purple.

Me: Again, the grass is not purple, the reason why you are seeing samples of purple grass is because it’s been modified, original grass is green.

You: Psychologists need to work with what they see, right now we see empirical evidence of purple grass that we can reproduce, so grass is purple ?

Me:

Edited April 18, 2022 by no_name

[Carl-Richard]

40 minutes ago, no_name said:

You are misusing terms again. Correlation measures relation between two variables. I am talking about proportions. 

Maybe you did there, but if we're indeed talking about the same typology, it's claimed that specific functions go together, i.e, correlate, like I stated earlier (e.g. Ti and Fe etc.). It's these functions that will occur in different proportions.

[Carl-Richard]

2 hours ago, no_name said:

Also, no offence, but talking to you feels a bit like this right now:

You: This wouldn’t work because the grass is purple, the experiments show that the grass is purple.

Me: The grass is green, it’s just that when you add certain chemical to the fertilizer the grass turns purple. 

You: Yea, the grass is purple and it’s the best what we’ve got so far, I see all this purple grass so I can verify that the grass is purple.

Me: Again, the grass is not purple, the reason why you are seeing samples of purple grass is because it’s been modified, original grass is green.

You: Psychologists need to work with what they see, right now we see empirical evidence of purple grass that we can reproduce, so grass is purple ?

Me:

Nooo I'm just interested in the empirical aspect of all this  I'm not making any statement like "cognitive functions ARE false" or "Big 5 ARE true". It seems like we're on the same grounds. It's just my prior misunderstandings (mixing up typology, MBTI, tests etc.) messed up the discussion.

[thisintegrated]

6 hours ago, no_name said:

So you don’t believe the +-+- system has any grounding to it? I don’t see how someone could be Ni and Ne or Ti and Te, in my head there exists a clear preference.

Also in your example there someone who has Ne and Ni as their first 3 functions, wouldn’t that person be some kind of hallucinating psycho? Don’t they need an S function to survive? 

As an Ni dominant I appreciate and understand Ne, but it gives me anxiety if I had to do it and I don’t like doing it. E.g., if you were to ask me to brainstorm, I would get tired of it and start clinging to some one item instead that I find most interesting, yet I do like it when someone does the brainstorming for me and shows me other options 

I do not understand why you can’t have an NiFiTiSe for example though, but I am still just learning so maybe there is a reason..

Preference vs strength is the difference.  Although Ni is my 5th function, it's stronger than the Ni of a sensor for whom it's their 3rd or 4th function.

I may default to Ne, but the more I use Ne the more I need Ni to balance it.  Ne without Ni is useless as you'll just be aware of potential solutions but will never actually find the solution.

 

Quote

So you don’t believe the +-+- system has any grounding to it?

I think there's some truth to it.  It seems to work, but so does Socionics despite working completely differently.

I'm still trying to determine how accurate it is.  It's really just a technique for manipulation so I don't use it normally.

 

Quote

Don’t they need an S function to survive? 

Everyone uses all 8 functions.  Just half of them are considered "unconscious" by MBTI.  Socionics focuses more on all 8 than MBTI.

 

Quote

As an Ni dominant I appreciate and understand Ne, but it gives me anxiety if I had to do it and I don’t like doing it. E.g., if you were to ask me to brainstorm, I would get tired of it and start clinging to some one item instead that I find most interesting, yet I do like it when someone does the brainstorming for me and shows me other options 

Hmm, that's interesting.  I love both Ne and Ni.  They should always work together.
 

Quote

I do not understand why you can’t have an NiFiTiSe for example though, but I am still just learning so maybe there is a reason..

It seems "neat" or "logical" that you'd have an extroverted function after every introverted function, so some people take it as fact.  Maybe there's truth to it but, in any case, I think it's more useful to order functions by strength. 

 

 

 

 

Edited April 18, 2022 by thisintegrated

[something_else]

On 17/04/2022 at 0:32 AM, thisintegrated said:

And why did the algorithm fail at even telling apart extroverts from introverts?  This is a clear indication the study is inherently flawed.  

If I remember correctly, the slot it did best on by far was T vs F. The problem in this case was also that most of the users in the dataset were introverted intuitive because it's from an online forum, so it didn't have as many examples of extroverts to learn from. Whereas thinking vs feeling was split close to 50/50. And I'd say online the difference between thinking vs feeling is gonna be way more noticeable than introversion vs extroversion.

It wasn't a perfect study by any stretch, not even close. But IMO the fact not a single statistics expert could come up with a model that did very well is still worth noting. The conclusion was basically there is only very little statistical difference between the online speech patterns of different INXX types

Quote

Just letting AI "go at it" won't produce any meaningful information

How do you know? We taught AI to play chess and Go better than humans have learnt to play those games in thousands of years by literally just 'letting it go at it.' Took the AI 4 hours to be at a level better than any human has ever been.

It doesn't need to know anything about cognitive functions, it simply needs to find patterns that all Judging users in the dataset exhibit, or all Introverted users in the dataset exhibit. Then it can identify the other Introverted users or the other Judging users who have similar patterns. This is textbook text classification problem, but it only works if there is actually a correlation between how a person writes online and their MBTI type

Quote

Introversion/extroversion isn't arbitrary nonsense.  It's a very obvious and well-accepted phenomenon.  Your denial of this is clearly indication of your bias.

I think bias is also an issue for you. Totally ignoring something like the Big 5 simply because you think it's too difficult/different to use is a very culty thing to do. It's extremely similar to MBTI except you have slightly different categories and it's on a percentage scale instead of there only being binary options for each slot. Personality is complex and different models are better for doing different things. If you want to make the most accurate predictions about people's actions that are rigorously verifiable by others, Big 5 is the best model we have

However, it's purely statistically derived, so it lacks some of the deeper ideas and qualitative analysis that came from Jung, which cognitive functions let you make. I totally get that. They're also more fun to use and ponder. I get why you like the model. It's well designed in a lot of ways and it's 'deeper' than something like Big 5 which is harshly cold and logical

MBTI also has a habit of being a bit culty, though. I was obsessed with it for years, then eventually I realised I was pulling half of the analysis I was making out of my ass, I was making very grand deductions from very little data, viewing almost everything about a person through the lense of their MBTI type and function stack. You look at posts like this on the MBTI subreddit and it's just so fucking cringe, abstract, and divorced from reality, yet I distinctly remember when I used to think and talk like this:

https://www.reddit.com/r/mbti/comments/bemvod/8th_slot_ti_vs_7th_slot_ti/

Read that, and the comments. These people are just telling each other stories over and over until they all speak the same language and then pat each other on the back and call it an accurate analysis of personality. It's so culty

Edited April 18, 2022 by something_else

[thisintegrated]

35 minutes ago, something_else said:

How do you know? We taught AI to play chess and Go better than any human in the world in 4 hours by literally just 'letting it go at it'

It's just as likely to identify Spiral Dynamics stages as MBTI types unless you teach it what MBTI is.

To get good at chess, AI still had be told how the pieces move, and what the win condition is.  Chess pieces have arbitrary rules for movement.  AI would never "figure it out" by itself.  But once the AI understands the arbitrary rules, then it can start learning.

Cognitive functions is a more complex concept than how chess pieces can move, so making sure the AI understands them takes more time and effort.  Unlike with chess, you can't simply input the rules and let it run, as the rules themselves are complex.  For chess, you just say "x moves up, y moves up and to the left" and then you let it run through every combination, within these rules.  For MBTI, you'd need to provide gigabytes of data just for the rules.  But to exactly figure out the rules, you'd need a separate AI to train for every single function, which in itself is a big and ambitious project. 

People usually need at least a dozen examples (+a good explanation) of each function to really get a feel for it.  For an AI you'd need waaaaaay more examples to get it to even have a partial understanding of a function.  And if you can't teach it every function, that's like failing to explain the rules of chess to a chess-bot.  

Edited April 18, 2022 by thisintegrated

[Danioover9000]

    Seriously, when I learned about cognitive functions I felt like a smarter person. I understand now more that people really do have more complex personality types.

[thisintegrated]

Just now, Danioover9000 said:

    Seriously, when I learned about cognitive functions I felt like a smarter person. I understand now more that people really do have more complex personality types.

MBTI isn't personality types, but "Information metabolism".

[Carl-Richard]

On 18.4.2022 at 6:13 PM, something_else said:

MBTI also has a habit of being a bit culty, though. I was obsessed with it for years, then eventually I realised I was pulling half of the analysis I was making out of my ass, I was making very grand deductions from very little data, viewing almost everything about a person through the lense of their MBTI type and function stack. You look at posts like this on the MBTI subreddit and it's just so fucking cringe, abstract, and divorced from reality, yet I distinctly remember when I used to think and talk like this:

I've also gone through the same process, and it's one of the reasons why I've been so insistent on focusing on the empirical side of things here. The whole premise of applying the model to see if it makes sense is highly questionable. It relies on so many assumptions, it makes your head spin. For one thing, it relies on nothing but your own mind, which is a product of your genes, your environment and your unique life story. Even if similar people agree on something, their environment is often shared and limited by cultural biases. Your mind's existence is wholly dependent on self-interest, emotional desires and survival needs, and is prone to biases of belief, data selection, and flawed reasoning, memory, perception etc. Even the scientific enterprise with its multiple-party oversight and structured methodologies, struggles to keep all of this in check, so to assume that you yourself are transparent enough and responsible enough to hold yourself to account, well that's rich!

[thisintegrated]

38 minutes ago, Carl-Richard said:

It relies on so many assumptions, it makes your head spin. For one thing, it relies on nothing but your own mind

*On a time-proven system.

 

Here's the thing.  Even if MBTI was complete BS with each type being completely random, it would still work thanks to the fact that we have thousands of examples on personality database.  Get a list of ENTPs, go "AHA!", and now you know the general vibe of ENTPs!  Even a complete novice with a mental impairments will easily spot the difference between an ENTP and an ISFJ.  Again, MBTI just gives names to real phenomenon.  Instead of "ENTP & ISFJ" you could just say "cool unconventional nerd & traditional, moral parent".  But the MBTI system is obviously the better alternative.

 

38 minutes ago, Carl-Richard said:

it relies on nothing but your own mind, which is a product of your genes, your environment and your unique life story. Even if similar people agree on something, their environment is often shared and limited by cultural biases. 

No.  Some cultures may prefer "nerdy" or "sensor" or "success-driven" types more than others, sure.  But different cultures aren't gonna have different interpretations of Ti.  Ti has a clear definition and I highly doubt you'll find a culture that already has a concept of "Ti" in their society with strict definitions that differ from MBTI's.

 

38 minutes ago, Carl-Richard said:

Your mind's existence is wholly dependent on self-interest, emotional desires and survival needs, and is prone to biases of belief, data selection, and flawed reasoning, memory, perception etc.

Speak for yourself, man.  You're an affiliative type, and non-systematic thinker, so according to MBTI you would, indeed, have all of these problems you've just listed.  I, for example, am not at all biased in my typing.  What could intentionally mistyping someone (because you don't like them?) possibly accomplish?  If I hated someone and wanted to make them suffer, I'd be extra careful to type them accurately so I can manipulate them effectively.

 

38 minutes ago, Carl-Richard said:

Even the scientific enterprise with its multiple-party oversight and structured methodologies, struggles to keep all of this in check, so to assume that you yourself are transparent enough and responsible enough to hold yourself to account, well that's rich!

What you mean?  

[Carl-Richard]

1 hour ago, thisintegrated said:

*On a time-proven system.

- filtered through your own mind.

 

1 hour ago, thisintegrated said:

Here's the thing.  Even if MBTI was complete BS with each type being completely random, it would still work thanks to the fact that we have thousands of examples on personality database.  Get a list of ENTPs, go "AHA!", and now you know the general vibe of ENTPs!  Even a complete novice with a mental impairments will easily spot the difference between an ENTP and an ISFJ.  Again, MBTI just gives names to real phenomenon.  Instead of "ENTP & ISFJ" you could just say "cool unconventional nerd & traditional, moral parent".  But the MBTI system is obviously the better alternative.

It's a dodgy website where people don't even agree what an ENTP is. If 10/10 voted the same type but gave 10 incommensurable answers for why, then that tells you nothing.

 

1 hour ago, thisintegrated said:

No.  Some cultures may prefer "nerdy" or "sensor" or "success-driven" types more than others, sure.  But different cultures aren't gonna have different interpretations of Ti.  Ti has a clear definition and I highly doubt you'll find a culture that already has a concept of "Ti" in their society with strict definitions that differ from MBTI's.

But you don't know that, which is all I'm saying. Jung's clinical patients were not cross-cultural, and your experience (like Jung's) is not universal.

 

1 hour ago, thisintegrated said:

Speak for yourself, man.  You're an affiliative type, and non-systematic thinker, so according to MBTI you would, indeed, have all of these problems you've just listed.  I, for example, am not at all biased in my typing.  What could intentionally mistyping someone (because you don't like them?) possibly accomplish?  If I hated someone and wanted to make them suffer, I'd be extra careful to type them accurately so I can manipulate them effectively.

Cognitive biases are notorious for being involuntary and unconscious.

 

1 hour ago, thisintegrated said:

What you mean?  

Peer reviewed journals, statistical methods, standards of validity and reliability etc. Despite all that, there are still subpar articles being published every day (and excellent ones being rejected). It's not an unknown phenomena of scientists tinkering with the numbers to further their own careers. People are also invested in maintaining the current paradigm (their careers and worldviews). Even for the mightiest of geniuses, toppling that is an uphill battle.

[thisintegrated]

1 hour ago, Carl-Richard said:

It's a dodgy website where people don't even agree what an ENTP is. If 10/10 voted the same type but gave 10 incommensurable answers for why, then that tells you nothing.

The most reputable site of its kind.

Statistical accuracy is mostly dependent on sample size.  If 1000 people vote, then the probability of the top type being correct is incredibly high.

 

1 hour ago, Carl-Richard said:

If 10/10 voted the same type but gave 10 incommensurable answers for why, then that tells you nothing.

With a large enough sample size, and/or a basic understanding of the theory, this isn't a valid argument.

 

1 hour ago, Carl-Richard said:

But you don't know that, which is all I'm saying. Jung's clinical patients were not cross-cultural, and your experience (like Jung's) is not universal.

 

Cognitive biases are notorious for being involuntary and unconscious.

16personalities.com/country-profiles/global/world Have fun

 

No cultures seem to have any difficulty understanding MBTI.