pjeby

• pjeby 14 Nov 2019 3:53 UTC

  3 points

  in reply to: Mary Chernyshenko's comment on: The Curse Of The Counterfactual

  I’m not sure I un­der­stand your ques­tion. In or­der to think that there’s a prob­lem with how much love he’s pro­vid­ing, you have to have a coun­ter­fac­tual in which he’s sup­posed to be pro­vid­ing more. For the amount of love to be in­suffi­cient, there has to be some­thing to com­pare it to. If you aren’t (im­plic­itly) com­par­ing, then there is noth­ing to draw it to your at­ten­tion in the first place.

  In other words, you wouldn’t keep say­ing “I guess he didn’t”, be­cause if you’re not com­par­ing, then there’s not an is­sue any more—it’s just his­tory, not an un­re­solved prob­lem.

  It sounds to me like the ex­pe­rience you’re talk­ing about is in­com­plete grief, like maybe a de­scrip­tion of a situ­a­tion where some­one is ac­cept­ing (at least in­tel­lec­tu­ally) that they aren’t go­ing to get the love they want in the fu­ture, but has yet to ac­cept that they didn’t get it in the past. Be­cause as long as they think they should have got­ten it, the griev­ing is still in­com­plete.

  As for dead­en­ing, let­ting go of things gen­er­ally makes us more al­ive, not less, be­cause we stop ob­sess­ing over the things we can’t change, and move on to en­joy­ing what we have (or can ac­tu­ally get). But be­fore one ac­tu­ally lets go of some­thing, the idea of let­ting go feels like it would be a loss.

  As I sug­gested in the ar­ti­cle, our brains treat un­ac­knowl­edged losses like they are still as­sets on our in­ner books of util­ity. So the idea of writ­ing off a loss feels like it is a loss. But once the write-off is ac­tu­ally done, then it no longer feels like a loss, be­cause it’s now the sta­tus quo, and there­fore it doesn’t keep com­ing back to con­scious at­ten­tion the way a per­ceived threat of loss does.

• pjeby 9 Nov 2019 6:17 UTC

  3 points

  in reply to: henryaj's comment on: Im­ple­ment­ing an Idea-Man­age­ment System

  That just means you’ve not seen that many wikis. ;-) For ex­am­ple, the Con­nect­edText per­sonal wiki soft­ware in­cludes back­links, date-spe­cific pages, and graph vi­su­al­iza­tion of link struc­tures, much like Roam. It also has the abil­ity to in­clude pages in oth­ers, and some of Roam’s other fea­tures could likely be em­u­lated us­ing CT’s script­ing and tem­plat­ing sys­tems, though it’d be a pain.

  I ac­tu­ally own a copy of an older ver­sion of CT but stopped us­ing it many years ago be­cause it’s not ter­ribly in­ter­op­er­a­ble with any­thing else.

• pjeby 4 Nov 2019 16:01 UTC

  5 points

  in reply to: moses's comment on: Skill and leverage

  This pre­sup­poses that you know what the difficulty level is for the per­son in ques­tion. It also ig­nores a ton of stuff that can get be­tween “easy thing” and “ac­tual do­ing”, like what their pri­ori­ties, in­ter­ests, and abil­ities are.

  Let’s say Bob has a re­ally im­por­tant pro­ject he needs to work on. He’s stuck and ob­sessed with it. Mean­while, his room goes un­cleaned and his dish­washer un­loaded. He’s not ac­com­plish­ing any­thing, but he’s not do­ing those sim­ple things be­cause he’s pour­ing en­ergy into some­thing else.

  Now let’s con­sider Alice. Alice is a blind para­plegic com­puter pro­gram­mer, who runs rings around her peers when it comes to cod­ing. Pro­gram­ming for her is su­per easy, barely an in­con­ve­nience. But clean­ing up the room or load­ing the dish­washer are not ex­actly her strengths.

  And then there’s Carl. He spends hours play­ing video games at in­sanely high difficulty lev­els that no­body else can match. But putting away dishes is bor­ing, and doesn’t get him that sweet sweet cred… or en­dorse­ment deals and ad­ver­tis­ing rev­enue. He’ll do it to­mor­row, for sure. Maybe. Or maybe his mom will.

  None of these peo­ple’s rooms are get­ting cleaned or dish­wash­ers loaded, but that fact by it­self tells you very lit­tle about what that per­son can ac­com­plish. (After all, Bob could eas­ily be a suc­cess­ful best-sel­l­ing au­thor who lets his place go to hell when he gets stuck in the mid­dle of a book pro­ject.)

• pjeby 4 Nov 2019 15:49 UTC

  12 points

  in reply to: Kaj_Sotala's comment on: The Curse Of The Counterfactual

  Hey Kaj. I was ac­tu­ally look­ing for feed­back in email, but this is good too. :) (I’ll up­date the ar­ti­cle to clar­ify on that point.) Thanks for the info about your friend’s ex­pe­rience: the an­swer to their ques­tion is that the act of vi­su­al­iz­ing re­quires them to ac­cess im­plicit in­for­ma­tion from their mem­ory from di­rect (if re­mem­bered) ex­pe­rience, vs. sim­ply ver­bal­iz­ing cached facts. It is struc­turally similar to scan­ning one’s mem­ory for past ex­pe­riences, look­ing for some­thing that matches a pat­tern of feel­ing or be­hav­ior. I’m only us­ing the term “felt sense” be­cause there’s no sense (no pun in­tended) in cre­at­ing yet an­other name for some­thing that is already de­scribed in other places. (Also, some peo­ple ac­tu­ally do ac­cess the turn in­for­ma­tion kines­thet­i­cally, i.e., by feel­ing their way through the re­called day.)

  As to your tran­script, I see you tran­si­tioned from the Quick Ques­tions right to the Work, which is a good move in the event one ob­jects to one’s de­sires. But I think per­haps you’ve missed some­thing (two some­things, ac­tu­ally) about how the Work works.

  So, when you got to: “what hap­pens, when you be­lieve that thought?”, you took the re­sponse you got as an ob­jec­tion from a part (mix­ing IFS in), rather than sim­ply tak­ing the re­sponse at face value. In other words, “What hap­pens when I be­lieve this thought? I feel like the reins are pul­ling me to my death”. You ac­tu­ally got the an­swer to your ques­tion! When you be­lieve the thought that it’s im­pos­si­ble to do any­thing mean­ingful be­cause you’ll get pul­led, the con­se­quence is just that: feel­ing like you’re be­ing pul­led to death.

  The next ques­tion, “who would I be with­out that thought?” would then be helpful in tar­get­ing the spe­cific be­lief, be­cause ob­jec­tions to let­ting go of the be­lief di­rectly im­ply the state of the world (or your­self) that your be­liefs pre­dict would re­sult from you not be­liev­ing it.

  This might’ve avoided a lot of the go­ing in cir­cles you did from this point on in the tran­script, and led you di­rectly to the tar­get schema with less… well, thrash­ing be­tween ideas, for lack of a bet­ter word.

  The rea­son I’ve moved to­wards us­ing the Work as a prime in­ves­tiga­tive tool is that it lets you walk the be­lief net­work re­ally fast com­pared to other meth­ods. Get­ting your brain to ob­ject to get­ting rid of a be­lief forces it to re­veal what the next be­lief up the branch is with far less wasted move­ment.

  And as you can see, start­ing from a place where you already have a con­crete ob­jec­tion (e.g. us­ing a tool like the Quick Ques­tions), you can move re­ally rapidly to the real “meat” of an is­sue.

  That be­ing said, the Quick Ques­tions are de­signed to solve lo­gis­ti­cal prob­lems, more than emo­tional ones—aside from the emo­tional is­sue of fo­cus­ing on the prob­lems in­stead of on solu­tions. A Minute To Un­limit You is just a men­tal ju­jitsu move to dis­en­gage your brain’s plan­ning sys­tem from “There’s a Prob­lem” mode and put it back into “Seek­ing Solu­tions” mode.

  Of course, that’s only one mod­ule of your brain’s mo­ti­va­tion sys­tem, as the ebook men­tions. There are four other mod­ules (like the two that han­dle pun­ish­ment and virtue-sig­nal­ling) that can be in­volved in a mo­ti­va­tion prob­lem, but it’s usu­ally eas­iest to be­gin with the Quick Ques­tions to rule out a mode 1 mis­match first, even if the prob­lems be­ing pre­dicted turn out to be com­ing from one of the other mod­ules.

• pjeby 1 Nov 2019 23:43 UTC

  21 points

  in reply to: Ben Pace's comment on: The Curse Of The Counterfactual

  Hi Ben, thanks for com­ment­ing.

  What I’d first like to say is that nega­tive re­in­force­ment and pun­ish­ment are ac­tu­ally two differ­ent things. What you’re de­scribing as “pun­ish­ment” is ac­tu­ally just nega­tive feed­back: i.e. notic­ing that some­thing you’re do­ing isn’t work­ing. But pun­ish­ment is some­thing we do to raise some­one’s costs for bad ac­tion. This does not nec­es­sar­ily re­sult in any re­in­force­ment for the sub­ject of the pun­ish­ment.

  In “Ing­var’s” case, for ex­am­ple, he con­stantly pun­ished him­self for sur­fing the in­ter­net, but this was ac­tu­ally pos­i­tively re­in­forc­ing for the be­hav­ior of self-pun­ish­ment it­self, and did noth­ing to dis­cour­age the in­ter­net sur­fing be­hav­ior!

  Even within the tech­ni­cal con­text of be­hav­iorist learn­ing, “pun­ish” and “nega­tively re­in­force” are two differ­ent things… and pun­ish­ment does not do what you seem to be think­ing it does.

  Tech­ni­cally, what hap­pens when you pun­ish an an­i­mal or per­son, is that you end up pos­i­tively re­in­forc­ing what­ever works quick­est to stop the pun­ish­ment. Pu­n­ish­ment, in and of it­self, does not ac­tu­ally al­ter be­hav­ior. The only thing it trains you (or any other an­i­mal) to do is to avoid the pun­ish­ment.

  And when you are ap­ply­ing so­cial pun­ish­ment of the type de­scribed in this ar­ti­cle, the thing that stops it is (e.g. in Sara’s case) ideation. The thing that turns off self-pun­ish­ment is imag­in­ing a fu­ture in which you are a bet­ter per­son and the bad thing can’t hap­pen any more. So, in a be­hav­iorist re­in­force­ment sense, by pun­ish­ing your­self in this fash­ion you are train­ing your­self to imag­ine bet­ter fu­tures, be­cause that’s the fastest way to stop the pain.

  IOW, prop­erly un­der­stood, the only func­tional use of pun­ish­ment is to raise the costs of bad be­hav­ior. But in a self-ap­plied case, rais­ing your own costs is not a func­tional thing to do, es­pe­cially when you fac­tor in the moral li­cens­ing for be­ing vir­tu­ously self-pun­ish­ing, and effec­tively train­ing your­self to imag­ine things be­ing bet­ter, in­stead of ac­tu­ally do­ing any­thing to make them bet­ter.

  So in that sense, I will say, no, it’s not the case that pun­ish­ing your­self (us­ing ei­ther the so­cial or be­hav­iorist defi­ni­tion) is a use­ful strat­egy for any­thing other than con­vinc­ing oth­ers not to pun­ish you (worse) for the same thing. That is the one way in which pun­ish­ing your­self is ac­tu­ally use­ful, and it’s of­ten how we learned to do it. (That is, to pun­ish our­selves for the same things our par­ents pun­ished us for, to lessen their de­sire to pun­ish us.)

  That be­ing said, we prob­a­bly have differ­ent defi­ni­tions of what “pun­ish­ment” ac­tu­ally con­sists of. In this post, I mean in the sense of “at­tack­ing rep­u­ta­tion to raise the tar­get’s costs”, not “nega­tive feed­back to shape be­hav­ior”, which is some­thing else al­to­gether.

  Peo­ple rou­tinely con­fuse these two things, be­cause our moral bias tells us that we must not let wrongs go un­pun­ished. So we dis­tort what be­hav­iorism ac­tu­ally says about learn­ing into “re­ward and pun­ish­ment”, when in fact nei­ther re­ward nor pun­ish­ment are re­li­able re­in­force­ment strate­gies! (For one thing, re­wards and pun­ish­ments are usu­ally too far away in time from the ac­tual be­hav­ior to have any mean­ingful effect, though that’s not the only differ­ence.)

  The mind­set of re­in­forc­ing ac­tual be­hav­ior, vs. re­ward­ing and pun­ish­ing what we think should be done, are very, very differ­ent in prac­tice, but our brains are bi­ased to­wards con­fus­ing the two.

  As for Sara, I think per­haps you are over­gen­er­al­iz­ing from Car­los’s ex­am­ple. I have differ­ent ex­am­ples in the ar­ti­cle be­cause there are many differ­ent ways for “pun­ish­ing based on coun­ter­fac­tu­als” to man­i­fest. What I did not cover in Sara’s case (or Ing­var’s for that mat­ter) is that the sur­face-level “shoulds” be­ing dis­cussed were not the root is­sue. As I men­tion later in the ar­ti­cle, one be­gins with what­ever one is aware of, but work­ing on these ini­tial “should” state­ments then leads us deeper into the be­lief net­work.

  For ex­am­ple, Ing­var be­lieved he should have been work­ing, and should have been able to finish in a cer­tain amount of time. But the solu­tion to this prob­lem was not “grieve for not hav­ing worked”! It was dis­cov­er­ing that the real is­sue was be­liev­ing he was a bad per­son un­less he was work­ing. Re­mov­ing that be­lief stopped him from gen­er­at­ing coun­ter­fac­tu­als about how he should have been work­ing, which then led to him think­ing of ways to ac­tu­ally get the work done.

  IOW, it’s the de­ac­ti­va­tion of the pun­ish­ment sys­tem that’s rele­vant here, be­cause its ac­ti­va­tion blocked him from think­ing about the ac­tual pro­cess of work and the trade-offs in­volved, due to the “sa­cred­ness” of pun­ish­ing him­self for be­ing a lazy evil­doer who wasn’t work­ing.

  In the same way, Sara’s root is­sue isn’t that she’s pun­ish­ing her­self for her failed ac­tions, it’s that she be­lieves she needs to prove her­self… or else she’s not a ca­pa­ble per­son. It’s that un­der­ly­ing be­lief which mo­ti­vates the gen­er­a­tion of the coun­ter­fac­tu­als in the first place.

  The full chain of events (for Sara and Ing­var) looks some­thing like this:

  • Step 1: Learn that a per­sonal qual­ity or be­hav­ior is sub­ject to pun­ish­ment by oth­ers (e.g. bad­ness, in­com­pe­tence)

  • Step 2: Try to avoid feel­ing bad by cre­at­ing an ideal of some kind (e.g. pun­ish one’s self for evil, seek recog­ni­tion to prove com­pe­tence) that will coun­ter­act this and avoid fu­ture punishment

  • Step 3: En­counter situ­a­tions in life that re­mind one’s self of the qual­ity learned about in step 1

  • Step 4: Gen­er­ate coun­ter­fac­tu­als based on the ideal to stop the pun­ish­ment (Sara) or pun­ish one’s self for failing to make the ideal hap­pen (Ing­var)

  Here’s the thing: the only part of this cy­cle that you can mean­ingfully change is the learn­ing found in step 1, be­cause oth­er­wise ev­ery time they en­counter a re­minder in the world, the pun­ish­ment will be re­mem­bered, and sus­tain the mo­ti­va­tion for avoidance. Without this pun­ish­ment cy­cle in effect, the per­son can ac­tu­ally think about what would be a good way to reach their goals. But with the cy­cle in effect, all the per­son can think about when it comes up is what’s the fastest way to make the hurt­ing stop!

  I cov­ered this more with the Ing­var ex­am­ple than the Sara one, but know­ing how to do some­thing doesn’t help in this cy­cle, be­cause it pro­duces the “yeah, but...” re­sponse. From in­side of this cy­cle, prac­ti­cal ad­vice liter­ally seems ir­rele­vant or off-topic, or at best mis­guided. Peo­ple in­side the loop say things like, “yeah, but it’s not that sim­ple” or “you just don’t un­der­stand”, when you try to give them prac­ti­cal ad­vice.

  ISTM that you have over­gen­er­al­ized from Car­los’ ex­am­ple that this is pro­cess is all about grief. But even in Sara’s case, it’s im­por­tant to un­der­stand that she can­not ac­tu­ally ac­cept or act on nega­tive feed­back with­out first ac­knowl­edg­ing what ac­tu­ally hap­pened. If there’s a se­man­tic stop sign in her brain that pops up ev­ery time she tries to con­sider ways to be­have (be­cause in or­der to do that she has to think about what she ac­tu­ally did or might do), then she can’t re­ally think about how to act differ­ently, only ru­mi­nate about how she ought to have done some­thing else.

  So when we say “we should have done X” or “I should do Y”, we are not ac­tu­ally say­ing the full truth. What we are do­ing is deny­ing the un­der­ly­ing re­al­ity that we did not do X, and we don’t want to do Y.

  Sara ac­tu­ally knew, go­ing into the con­fer­ence, that she tended to be stub­born, and speci­fi­cally thought ahead of time that she should not be. The prob­lem is that “I should not do X” is an ar­gu­ment with re­al­ity: you know full well ahead of time that you prob­a­bly will do X, but see this as wrong (in a moral sense, rather than a func­tional one). This mo­ti­vates you to deflect the per­cep­tion (and as­so­ci­ated pun­ish­ment) by as­sert­ing that you should do the right thing. (Like Ing­var as­sert­ing he should get the work done in an af­ter­noon.)

  I hope that the above ex­pla­na­tion clar­ifies bet­ter what this ar­ti­cle is driv­ing at. The is­sue is that any­time we start think­ing about what we or other peo­ple “ought” to do—as a moral judg­ment—we im­me­di­ately “taboo trade­offs” and dis­en­gage from prac­ti­cal rea­son­ing. We’re no longer in a state of mind where feed­back from what ac­tu­ally hap­pened is even be­ing taken into ac­count, let alone learned from.

  Fi­nally, as for your com­ments on re­la­tion­ships, I’m just go­ing to say that most of what you said has no real bear­ing on Car­los’s ac­tual situ­a­tion, which I will not com­ment fur­ther on as it would re­duce his anonymity. But I do want to ad­dress this point:

  I read the sec­tion on Car­los, and it seems like the ex­plicit con­tent was that you should always give up on re­la­tion­ships when they’re mak­ing you an­gry, and while there’s a deep truth to that with long-term re­la­tion­ships, I don’t think it should be the stan­dard the solu­tion. The stan­dard solu­tion to be­ing an­gry at some­one is to fol­low-through and make sure the cause is re­solved, such that your anger reaches its nat­u­ral con­clu­sion. This is true even when it’s built up for a while. Often there’s some­thing im­por­tant that’s been left un­said, and needs com­mu­ni­cat­ing.

  So, this is an over­gen­er­al­iza­tion, again, be­cause noth­ing in this post recom­mends any ob­ject-level be­hav­iors. What the post dis­cusses is the fact that, when you are coun­ter­fac­tu­al­iz­ing with moral judg­ment at­tached, you can­not rea­son prop­erly. Your brain hi­jacks your rea­son­ing in the ser­vice of your moral judg­ment, so you have liter­ally no idea what ac­tu­ally should be done on the ob­ject level of the situ­a­tion.

  The solu­tion to this prob­lem, then, is to dis­able the hi­jacker so you can get back in the cock­pit of the plane and figure out where you want to fly. In Ing­var’s case, he im­me­di­ately be­gan see­ing other ways he could be­have that would get to his goals bet­ter, and I had no need to ad­vise him on the ob­ject level. The is­sue was that with his moral judg­ment sys­tem ac­tive, he liter­ally could not even con­sider those op­tions se­ri­ously, be­cause they weren’t “pun­ish some­one” or “make the pain go away NOW”.

  With re­gard to re­la­tion­ships, as with ev­ery­thing else this ar­ti­cle talks about, the solu­tion is to be­gin with what­ever the ac­tual ground truth of the situ­a­tion is. If you are in­sist­ing that the other per­son in a re­la­tion­ship “should” be do­ing some­thing, and that the only solu­tion is to ex­press anger in their di­rec­tion, then you will miss the clue that some­times, be­ing an­gry at peo­ple doesn’t change them… but pos­i­tive re­in­force­ment might.

  (But of course, when we’re think­ing morally rather than strate­gi­cally, we think it’s wrong to use pos­i­tive re­in­force­ment, be­cause the other per­son doesn’t “de­serve” it. They should just do the right thing with­out be­ing re­warded, and they should be pun­ished for not do­ing the right thing. So saith the moral judg­ment brain, so shall it be!)

  Another prob­lem is where you say, “make sure the cause is re­solved, such that your anger reaches its nat­u­ral con­clu­sion”. The thing is, our anger’s “nat­u­ral con­clu­sion” is when some­body has suffered enough. (No­tice, for ex­am­ple, how some­body who ac­cedes to an­gry de­mands, but does not ap­pear re­morse­ful, will of­ten re­sult in the de­man­der get­ting an­grier. If it were about re­solv­ing the ac­tual is­sue, this would not make sense.) And suffer­ing enough doesn’t always cor­re­spond with an ac­tual solu­tion, ei­ther: note how of­ten peo­ple end up stuck in abu­sive re­la­tion­ships be­cause the abuser is re­ally good at ap­pear­ing re­morse­ful!

  So, fol­low­ing anger to its “nat­u­ral con­clu­sion” can eas­ily lead you astray, com­pared to clear­ing your head and act­ing strate­gi­cally. It can be al­most im­pos­si­ble to en­act, say, “tough love”, when you are stuck in your own mor­al­iz­ing about how some­one ought to be­have, both be­cause you can’t think it through, and be­cause it’s hard to do the “love” part while your brain is urg­ing you to make some­one to suffer for their sins.

  Any­way, in sum­mary: if you are ar­gu­ing ob­ject-level recom­men­da­tions from this ar­ti­cle, you’ve con­fused your in­fer­ences with my state­ments. The only ad­vice this post ac­tu­ally gives is to dis­en­gage your moral judg­ment if you want to be able to ac­tu­ally solve your prob­lems, in­stead of just ru­mi­nat­ing about them or pun­ish­ing your­self for them. (And I guess, to avoid re­cur­sively mak­ing a “should” out of this idea, since that’s just do­ing more of the prob­lem!)

  [Edit to add: I have added a new sec­tion to the ar­ti­cle, called The Dis­claimer, to clar­ify that none of the sto­ries con­tain, nor are in­tended to im­ply, any ob­ject-level ad­vice for the de­picted situ­a­tions, and that rather, the ar­ti­cle’s fo­cus is on the prob­lem of moral judg­ment im­pairing our abil­ity to rea­son about the truth, and even per­ceive what it is in the first place.]

The Curse Of The Counterfactual

• pjeby 1 Nov 2019 4:08 UTC

  5 points

  in reply to: Stefan_Schubert's comment on: Is there a defini­tive in­tro to pun­ish­ing non-pun­ish­ers?

  Be­fore think­ing of how to pre­sent this idea, I would study care­fully whether it’s true.

  I’m prob­a­bly refer­ring to the idea in a much nar­rower con­text, speci­fi­cally our in­cli­na­tion to ex­press out­rage (or even just mild dis­ap­proval) as a form of low-cost, low-risk so­cial pun­ish­ment, and for that in­cli­na­tion to ap­ply just as well to peo­ple who ap­pear in­suffi­ciently dis­ap­prov­ing or out­raged.

  The tar­gets of this in­cli­na­tion may vary cul­turally, and it might be an ar­ti­fact or side-effect of the hard­ware, but I’d be sur­prised if there were so­cieties where noth­ing was ever a sub­ject that peo­ple dis­ap­proved of other peo­ple not be­ing dis­ap­prov­ing of. Dis­ap­prov­ing of the same things is a big part of what draws so­cieties to­gether in the first place, so failing to dis­ap­prove of the com­mon en­emy seems like some­thing that au­to­mat­i­cally makes you “prob­a­bly the en­emy”.

  (But my rea­sons and ev­i­dence for think­ing this way will prob­a­bly be clearer in the ac­tual ar­ti­cle, as it’s about pat­terns of mo­ti­vated rea­son­ing that seem to re­li­ably pop up in cer­tain cir­cum­stances… but then again my ex­am­ples are not ter­ribly di­verse, cul­turally speak­ing.)

• pjeby 31 Oct 2019 21:51 UTC

  4 points

  in reply to: Ben Pace's comment on: Is there a defini­tive in­tro to pun­ish­ing non-pun­ish­ers?

  Yeah, those are the things I found, but none of them are the thing I re­mem­ber, which was some­thing that ex­plained how pun­ish­ment is costly (risky) for the pun­isher due to free-rid­ing by non-pun­ish­ers, so we evolved the de­sire to pun­ish non-pun­ish­ers in or­der to en­sure no­body gets away with free-rid­ing. None of these ar­ti­cles cover that, which is a sur­prise to me since I had to have read that idea some­where, and it feels to me like some­thing that’s part of the ra­tio­nal­sphere zeit­geist, yet I can’t seem to place where I ac­tu­ally read it.

  Any­way, I’m think­ing what I’ll do for now is link to this ques­tion from my ar­ti­cle, so peo­ple can see all the col­lected an­swers here. ;-)

[Question] Is there a defini­tive in­tro to pun­ish­ing non-pun­ish­ers?

• pjeby 31 Oct 2019 19:21 UTC

  8 points

  in reply to: Kaj_Sotala's comment on: On In­ter­nal Fam­ily Sys­tems and multi-agent minds: a re­ply to PJ Eby

  See, now this com­ment would have made a great ar­ti­cle. ;-) I think it says more clearly what you mean than the ar­ti­cle you ac­tu­ally wrote, and makes a much bet­ter case for your po­si­tion.

• pjeby 31 Oct 2019 19:15 UTC

  8 points

  in reply to: Kaj_Sotala's comment on: On In­ter­nal Fam­ily Sys­tems and multi-agent minds: a re­ply to PJ Eby

  Thank you for the con­sid­er­a­tion, and I ap­pre­ci­ate the ed­its. This was just an un­for­tu­nate con­fluence of events and I’m not hold­ing any grudges.

  I have to ad­mit that one of my faults is a healthy dose of the illu­sion of trans­parency. I tend to as­sume that other peo­ple can reach the same con­clu­sions I have when they have ac­cess to the same in­for­ma­tion my con­clu­sions are based on… even though there’s a dis­tinc­tion be­tween say, read­ing UtEB and grokking what it means about “legacy” ap­proaches to ther­apy.

  So some of the things you said in this ar­ti­cle seemed to me like ex­ces­sively be­la­bor­ing points I thought were already made quite ex­plic­itly in the text of UtEB, so I in­ter­preted it as you try­ing to ar­gue in fa­vor of IFS, not that you were just now re­al­iz­ing how IFS fit within UtEB’s model.

  The fact you posted an ar­ti­cle about UtEB be­fore made me as­sume that you un­der­stood it at least as well as I did (since in effect, you in­tro­duced me to it!), so I didn’t see why you would only be now dis­cov­er­ing those points… es­pe­cially since I thought they’d been cov­ered by our pre­vi­ous dis­cus­sion and your restate­ment of my po­si­tion.

  Re­gard­ing Core Trans­for­ma­tion, I’m glad you found it use­ful. Back at the time I men­tioned it, it was one of the bet­ter tech­niques available to me, de­spite the ten­dency to some­times get bogged down in “is that re­ally a part or am I imag­in­ing things” or parts get­ting in cir­cu­lar ar­gu­ments about things. But I later found that there were sim­pler ways to ad­dress the same things, be­cause what CT calls “core states” are also ac­cessible by sim­ply not ac­ti­vat­ing the parts of the brain that shut off those states. (e.g. by tel­ling us we don’t de­serve love)

  So if, for ex­am­ple, we don’t see our­selves as worth­less, then ex­pe­rienc­ing our­selves as “be­ing” or love or okay­ness is a nat­u­ral, au­to­matic con­se­quence. Thus I ended up purs­ing meth­ods that let us switch off the nega­tives and deal di­rectly with what CT and IFS rep­re­sent as ob­ject­ing parts, since these ob­jec­tions are the con­straint on us ac­cess­ing CT’s “core states” or IFS’s self-lead­er­ship and self-com­pas­sion.

  In effect, you could think of the ap­proaches I’ve been pur­su­ing since then as short­cut­ting the pro­cess of CT by jump­ing as di­rectly as pos­si­ble to our ob­jec­tions to ex­pe­rienc­ing our­selves as lov­able, okay, etc., and work­ing back­wards from there.

  To put it in con­text of your changes us­ing Trans­form­ing The Self, the shadow qual­ities (or “nega­tive qual­ities” as TTS calls them), are the things I tar­get first, since around 2012 or so.

  That’s be­cause prac­ti­cal ex­pe­rience had shown by then that al­most any­thing I tried to change in my­self or oth­ers us­ing other meth­ods would of­ten re­turn in a few weeks, un­less said nega­tive qual­ities were some­how ad­dressed. So, strate­gi­cally, go­ing hunt­ing for them first makes things a lot more effi­cient, as you then don’t have to worry about all the tac­ti­cal-level be­hav­iors and be­liefs be­ing re­gen­er­ated from the per­sis­tent, strate­gic-level, nega­tive self-image.

  In­ter­est­ingly, now that you’ve men­tioned TTS (in­di­rectly, by link­ing to your posts refer­enc­ing it), it re­minds me that TTS ac­tu­ally in­cludes some­thing rather like a re­con­soli­da­tion-ori­ented ap­proach to qual­ity changes. It might be in­ter­est­ing now to go back and re-read it with our newer knowl­edge of re­con­soli­da­tion in mind, to see if I can ei­ther im­prove on his tech­nique, or use some­thing from it to im­prove on mine.

• pjeby 31 Oct 2019 3:15 UTC

  11 points

  in reply to: Ben Pace's comment on: On In­ter­nal Fam­ily Sys­tems and multi-agent minds: a re­ply to PJ Eby

  I’m sorry you had a jar­ring ex­pe­rience be­ing named in the OP.

  Thank you. It is at least good to know that it was not his de­ci­sion to put this on the front page, though the num­ber of times I’m named still makes it feel a bit like it’s a call­ing out, es­pe­cially since he refers to “pje­byan” prac­tices as if they were what we dis­cussed, rather than the ma­te­rial from UtEB that he him­self pre­vi­ously posted.

  A lot of what we talked about in the origi­nal com­ments was ac­tu­ally what UtEB de­scribes as re­con­soli­da­tion, not what I do, be­cause I speci­fi­cally did not want to get into that here.

  Rather, my di­rect dis­cus­sion with Kaj was strictly fo­cused on the re­duc­tion­ism is­sue with parts-ori­ented mod­els, and the differ­ence be­tween de­liber­ate re­con­soli­da­tion (ala UtEB) and ac­ci­den­tal re­con­soli­da­tion (ala IFS). It was never sup­posed to be a refer­en­dum on my ap­proach to work­ing with clients or com­par­ing my ap­proach with IFS, out­side of me men­tion­ing some rea­sons why I don’t like to use parts-ori­ented ap­proaches (like IFS or any of its many pre­de­ces­sors), and how my ex­pe­riences re­late to what’s said in UtEB.

  In­deed, the only rea­son I felt safe to dis­cuss what I did in that pre­vi­ous thread was be­cause I could use UtEB as an ex­am­ple of a re­con­soli­da­tion-ori­ented ap­proach other than mine, be­cause I did not wish to cre­ate an im­pres­sion of us­ing LW as a pulpit from which to preach my own gospel. The un­ex­pected com­bi­na­tion of “sud­dently front­page” and “nam­ing names/​as­cribing po­si­tions” was quite un­pleas­ant, as it made it feel like I was be­ing shoved into a frame of do­ing that in di­rect op­po­si­tion to my at­tempts to keep the pre­vi­ous dis­cus­sion fo­cused on gen­eral schools of thought (e.g. be­hav­iorism vs. “parts”, de­liber­ate vs. ac­ci­den­tal re­con­sol­di­a­tion, etc.) rather than be­ing about “my way is bet­ter than yours”.

  After all, as the guidelines say, “aim to ex­plain, not per­suade”.

  (That be­ing said, I can also see how the frame shift prob­a­bly seems way more visi­ble and salient to me than it does to any­body else, and on a re-read of the ar­ti­cle, even I can see that the parts that got me up­set are re­ally very tiny in com­par­i­son to the whole. It’s also pretty un­der­stand­able in ret­ro­spect why Kaj could eas­ily have thought I was ar­gu­ing for a model of my own, rather than speak­ing gener­i­cally, with­out him hav­ing any in­ten­tion to dis­tort my views or at­tribute his own views to me… even as it’s also un­der­stand­able why the situ­a­tion in­clined me to give more weight to the re­verse hy­poth­e­sis.)

• pjeby 31 Oct 2019 2:58 UTC

  5 points

  in reply to: 9eB1's comment on: On In­ter­nal Fam­ily Sys­tems and multi-agent minds: a re­ply to PJ Eby

  My visceral re­ac­tion upon read­ing your com­ment above this one, for ex­am­ple, was that you were try­ing to de­mote IFS be­cause it sounds like you make a liv­ing pro­mot­ing this other non-IFS ap­proach.

  That fram­ing is ac­tu­ally part of what up­set me about this ar­ti­cle: it pre­sents some of my ar­gu­ments in a con­text that makes them seem as though they were made in sup­port of my own ap­proach vs IFS, rather than com­par­ing and con­trast­ing the ma­te­rial dis­cussed by two of Kaj’s own posts.

  In one post, he pre­sented re­con­soli­da­tion-ori­ented ther­apy as de­scribed in Un­lock­ing the Emo­tional Brain (UtEB for short), and in the other he dis­cussed IFS. My com­ments in the pre­vi­ous thread were about how UtEB’s ar­gu­ments re­gard­ing re­con­soli­da­tion show­case why IFS is an “ac­ci­den­tal re­con­soli­da­tion” model, and how a de­liber­ate model is more effi­cient. (Us­ing oc­ca­sional ex­am­ples from my ex­pe­riences with both types of ap­proach.)

  This post seems (to me at least) to frame that prior dis­cus­sion as if I was in­stead ar­gu­ing for my method­ol­ogy vs. IFS, when I was al­most ex­clu­sively ar­gu­ing “de­liber­ate vs. ac­ci­den­tal re­con­soli­da­tion”, with UtEB from Kaj’s own post as an ex­am­ple of the former va­ri­ety.

  So taken out of con­text, this post makes it sound as if I were do­ing just what you say: de­mot­ing IFS to pro­mote my own ap­proach. But the origi­nal con­ver­sa­tion was ac­tu­ally com­par­ing two schools of thought that Kaj had writ­ten ar­ti­cles about, and by ex­ten­sion, other schools that di­vide along the same lines.

  (But then, my view might be more than a lit­tle bi­ased by the un­ex­pected ap­pear­ance on the front­page, while think­ing that said ap­pear­ance was Kaj’s choice rather than a mod­er­a­tor’s, mak­ing me look ex­tra-close for why he made a choice that he didn’t ac­tu­ally make.)

• pjeby 31 Oct 2019 1:58 UTC

  8 points

  in reply to: Vaniver's comment on: On In­ter­nal Fam­ily Sys­tems and multi-agent minds: a re­ply to PJ Eby

  I came away think­ing that you think pjeby is right but some part of IFS is worth sal­vaging, and when I put my pjeby hat on, I can’t figure out which part you think is worth sal­vaging.

  Thank you, yes. That is ba­si­cally my po­si­tion as well, though it is also wrapped in a shell of “WTF am I be­ing named in the ti­tle and through­out the body of this post when it’s ac­tu­ally about Kaj’s po­si­tion and ap­pears only tan­gen­tially re­lated to what we talked about be­fore, be­cause al­though I did an­swer some ques­tions peo­ple asked about what I do, when I was speak­ing to Kaj I was mainly talk­ing about the dis­tinc­tions available be­tween IFS and UtEB, not dis­tinc­tions be­tween IFS and what I, per­son­ally, do with clients.” So this not only feels like a weird and con­fus­ing post on the level you ex­plain, it also feels like an at­tack on a straw­man ver­sion of what I do, be­cause Kaj ap­pears in some places to have con­fused my generic dis­cus­sion of “de­liber­ate vs. ac­ci­den­tal re­con­soli­da­tion” with me say­ing some­thing about my own, per­sonal method­ol­ogy.

• pjeby 30 Oct 2019 21:01 UTC

  15 points

  in reply to: Kaj_Sotala's comment on: On In­ter­nal Fam­ily Sys­tems and multi-agent minds: a re­ply to PJ Eby

  Mostly, I’m con­fused as to why this is a post at all (es­pe­cially front­page); it reads as an at­tempt to defend IFS against my crit­i­cisms of it, with­out ac­tu­ally dis­agree­ing with or re­fut­ing any of those crit­i­cisms. In which case, why post it at all?

  You seemed to have crit­i­cisms about IFS which looked to me like they were based on mi­s­un­der­stand­ings of IFS, so I tried to cor­rect them; and you seemed con­fused about why peo­ple found IFS valuable, so I tried to share my per­spec­tive on why peo­ple do.

  I’m still try­ing to find out where you think I have mi­s­un­der­stood IFS. Your ex­pla­na­tion of what you found valuable was use­ful, though, in that I can see the spe­cific in­sights or ex­pe­riences you are cred­it­ing to IFS that I was blind to be­cause I con­sider them min­i­mum re­quire­ments for a func­tional sys­tem, rather than bonus fea­tures of IFS in par­tic­u­lar.

  But that was largely an­swered in the com­ments of the pre­vi­ous thread, so as I said, I’m kind of con­fused by a front­page post that seems to be po­si­tion­ing it­self as a re­fu­ta­tion of my crit­i­cisms, but isn’t.

  I feel like I came to this con­ver­sa­tion mostly in a gears-ori­ented frame, where I don’t have any very strong agenda

  Me too… to the origi­nal con­ver­sa­tion. This post seems like an odd es­ca­la­tion on your part, to move from com­ments to a front­page post speci­fi­cally nam­ing me and mak­ing it less a quiet con­ver­sa­tion be­tween you and I to a pub­lic dis­pute wor­thy of front­page at­ten­tion. It’s par­tic­u­larly puz­zling be­cause the post spends a huge amount of time ve­he­mently.… not quite dis­agree­ing, but dis­agree­ing enough that a ca­sual ob­server might think, “Oh, Kaj re­ally gave PJ’s ideas a smack­down”, and in­ter­pret the situ­a­tion through a dom­i­nance frame, even though in ac­tu­al­ity you’ve barely dis­agreed at all. (In which case again, why post it? What new in­for­ma­tion does it ac­tu­ally add?)

  So a lot of what you’re see­ing from me is puz­zle­ment. If you were ac­tu­ally try­ing to in­form me of a mi­s­un­der­stand­ing, this doesn’t seem like a good way to go about it. But on the other hand, you haven’t ac­tu­ally in­formed me of any mi­s­un­der­stand­ings in this post, and in­stead pre­sented what looks like a weak pro­mo­tion of IFS and some fully-gen­eral coun­ter­ar­gu­ments. In which case, why bring me into it in par­tic­u­lar?

  I hon­estly can’t make heads or tails of why this is on the front page and not ei­ther con­tinued com­ments on the other thread, or a pri­vate con­ver­sa­tion, as it doesn’t seem to make sense to me no mat­ter how I look at it, based on the in­ten­tions you’ve stated. In­stead, it looks like an un­nec­es­sary (and pos­si­bly hos­tile) es­ca­la­tion. Rather than im­pute such in­ten­tions to you, my com­ments have been more try­ing to figure out what it is you were think­ing, us­ing my best guess for the ap­par­ent in­ten­tion of the post (i.e. pro­mote IFS/​defend it against my crit­i­cism).

  The idea that this post was to fur­ther mu­tual un­der­stand­ing is kind of be­yond my com­pre­hen­sion, since if that were the case I would’ve ex­pected you to ask ques­tions in the post rather than sim­ply stat­ing your case. (Or in some parts, just as­sum­ing your case rather than even stat­ing it!)

• pjeby 30 Oct 2019 17:16 UTC

  3 points

  in reply to: Stuart Anderson's comment on: What are some un­pop­u­lar (non-nor­ma­tive) opinions that you hold?

  Since your re­ply to my bring­ing up prob­lems is to bring up more stuff that has prob­lems, while not ad­dress­ing most of the prob­lems I pre­vi­ously raised, I don’t see how a con­ver­sa­tion can mean­ingfully pro­ceed from here, with­out it feel­ing like a Gish Gal­lop. For ex­am­ple:

  I cite the OKCupid data that speci­fi­cally sup­ports my state­ment here. I am say­ing men, as a class, do not favour a minor­ity of women.

  I read the ar­ti­cle you linked, and it says that 2 out of 3 mes­sages sent by men are to the women in the top ¹⁄₃ of at­trac­tive­ness, while on the other hand, women rate 80% of men as be­low-av­er­age at­trac­tive­ness… and then mes­sage most of them any­way.

  This sounds to me like it 100% con­tra­dicts your state­ments about men and women’s mat­ing prefer­ences.

  If the very data you cite liter­ally con­tra­dicts the premises you’re cit­ing it to sup­port, I don’t see how to have a sane con­ver­sa­tion about this, given that you don’t even re­motely touch on the ma­jor­ity of my ob­jec­tions. Also, the part where we have thor­oughly differ­ent value sys­tems means that there’s a ton of differ­ence in what’s con­sid­ered even rele­vant, so a mean­ingful dis­cus­sion is prob­a­bly not pos­si­ble.

• pjeby 30 Oct 2019 16:40 UTC

  8 points

  in reply to: Vaniver's comment on: On In­ter­nal Fam­ily Sys­tems and multi-agent minds: a re­ply to PJ Eby

  But this sort of re­minds me of my long-stand­ing at­tempt to figure out what Fris­ton-style mod­els are do­ing that PCT-style mod­els aren’t (and vice versa), where ac­tu­ally the thing at play seems to be a ques­tion of “which un­der­ly­ing the­ory de­serves more sta­tus?” than “what differ­ent pre­dic­tions do these mod­els make?”, in a way that’s quite difficult to re­solve.

  As far as differ­ing pre­dic­tions go, if IFS’ model of how parts need to be treated is cor­rect, then it should not be pos­si­ble to di­rectly al­ter some­one’s rules with­out any refer­ence to parts. Yet, the fact that non-parts-ori­ented ther­a­pies also work is a triv­ial re­fu­ta­tion of this pre­dic­tion. Ergo, the idea of parts is an un­nec­es­sary ad­di­tion from a the­o­ret­i­cal point of view.

• pjeby 30 Oct 2019 16:26 UTC

  3 points

  in reply to: Kaj_Sotala's comment on: On In­ter­nal Fam­ily Sys­tems and multi-agent minds: a re­ply to PJ Eby

  Eh, I think that dis­solv­ing agency is ac­tu­ally pretty darn im­por­tant, in a va­ri­ety of ways, both prac­ti­cal and the­o­ret­i­cal. But I think your at­tempts to sal­vage the idea of agency are tied to you see­ing it as es­sen­tial to the “pos­i­tive in­ten­tion” frame you see as part of IFS’ main ap­peal or value.

  So, I’ve writ­ten a sep­a­rate com­ment to ad­dress that, and the other prac­ti­cal-side ar­gu­ments of the ar­ti­cle.

• pjeby 30 Oct 2019 16:22 UTC

  11 points

  on: On In­ter­nal Fam­ily Sys­tems and multi-agent minds: a re­ply to PJ Eby

  Hav­ing ad­dressed the re­duc­tion­ist side (where you ba­si­cally agree with me, yet still think “agency” is use­ful), let’s now look at the prac­ti­cal­ities.

  As far as I can tell, your prac­ti­cal ar­gu­ment in fa­vor of IFS has three main parts:

  • Meta-schema (e.g. of seek­ing to ex­ter­mi­nate bad parts/​rules) are a thing that exists

  • IFS can ar­rive at the same place as other mod­els by differ­ent steps

  • IFS’ frame of pos­i­tive in­ten­tion is use­ful, and isn’t available in other models

  But none of these ar­gu­ments ac­tu­ally fa­vor IFS over other sys­tems, or even dis­t­in­guish it at all!

  For the first ar­gu­ment, the thing I pre­vi­ously de­scribed as “The In­ter­dict of Mer­lin for Self-Help” pre­vents peo­ple with meta-is­sues from re­solv­ing them with­out ei­ther 1) bang­ing their head on a wall till they no­tice there’s a wall there, or 2) hav­ing an­other liv­ing mind say, “hey, did you know that’s a wall you’re bang­ing your head against?”. In short, meta-is­sues are a Fully Gen­eral Coun­ter­ar­gu­ment be­cause they ap­ply to any method of change for hu­man brains. So they don’t dis­t­in­guish IFS from other sys­tems.

  Se­cond, of course IFS can ar­rive at the same place as other mod­els; my ar­gu­ment is that it lacks both rigor and sim­plic­ity in how it ar­rives at those places com­pared to mod­els based on pas­sive schema, that use an un­der­stand­ing of re­con­soli­da­tion to ex­plic­itly fo­cus on con­tra­dict­ing ex­plic­itly-se­lected mem­ory tar­gets.

  And the au­thors of UtEB speci­fi­cally note that ther­a­pies such as IFS work—to the ex­tent that they do—be­cause they are do­ing things that un­in­ten­tion­ally or in­di­rectly re­sult in re­con­soli­da­tion. So this sec­ond ar­gu­ment doesn’t ac­tu­ally dis­t­in­guish IFS from any other ther­apy that ac­ci­den­tally (i.e., with­out ex­plicit in­ten­tion/​tar­get­ing) pro­duces re­con­soli­da­tion.

  Fi­nally, the third ar­gu­ment is, I think, where the real thrust of this en­tire ar­ti­cle lies. I get the im­pres­sion in fact, that this third ar­gu­ment is the rea­son you’re defend­ing IFS so strongly in the first place: that this frame has had per­sonal mean­ing and util­ity to you.

  But that “pos­i­tive in­ten­tion” frame isn’t new, and definitely isn’t unique to IFS. Other ther­apy modal­ities had it long be­fore IFS was de­vel­oped, and you can use a re­duc­tion­ist modal­ity with­out los­ing the benefits.

  For ex­am­ple, I might say to some­one that some­thing they learned to was the best thing they could do in the situ­a­tion they were in, and back that up by helping them ex­pe­rience how and why that was the best thing, lead­ing to the same sort of sub­jec­tive ex­pe­rience as you de­scribe, of re­al­iz­ing that your brain is not, in fact, out to get you.

  And this fram­ing doesn’t re­quire me to pos­tu­late a part that has your well-be­ing in mind, yet for some rea­son keeps do­ing the same thing over and over! “You learned to do this be­cause it was good then, and it’s not so good now” is IMO a sim­pler the­o­ret­i­cal frame than “you have parts that are well-in­ten­tioned, but also kind of dumb”.

  (Of course, the way I do this, there’s an even more ba­sic ex­pe­rience, of one’s rules be­ing “think X, feel Y au­to­mat­i­cally”. By the time we’re talk­ing about any­thing like the pos­i­tive in­ten­tion frame, it’s already been es­tab­lished by sim­ple rep­e­ti­tion that the trou­bling thing acts like a rule that wires a but­ton di­rectly to a feel­ing. So the ques­tion raised is not, “why is this part do­ing this to me?” but “how did I learn that rule?”)

  So the third ar­gu­ment doesn’t dis­t­in­guish IFS from other meth­ods, and the sub­jec­tive ex­pe­rience of un­blend­ing via pos­i­tive in­ten­tion can be ar­rived at by other means, that IMO are more effi­cient as well as more epistem­i­cally cor­rect, and al­low for greater rigor and speed.

  IOW, on none of these three di­men­sions is IFS sub­stan­tially differ­ent from any “av­er­age ther­apy brand X”, with the pos­si­ble ex­cep­tion of its metaphor hav­ing in­tu­itive ap­peal for a self-helping user. Nor do any of the claimed benefits of a “parts”-fo­cused model not ap­ply equally to ap­proaches based on rules and learn­ing.

• pjeby 30 Oct 2019 7:59 UTC

  12 points

  in reply to: moridinamael's comment on: On In­ter­nal Fam­ily Sys­tems and multi-agent minds: a re­ply to PJ Eby

  Ex­am­ples of phe­nom­ena that con­tra­dict IFS model would be even more use­ful, though I’m failing to think of what those would look like.

  The rea­son it’s hard to think of what it would look like is that view­ing things through an agen­tic lens makes you miss those coun­terex­am­ples by con­fab­u­la­tion. For ex­am­ple, you can triv­ially ex­plain any be­hav­ior by pos­tu­lat­ing a part that wants to do that be­hav­ior. In con­trast, sim­pler mod­els have to be grounded in some­thing more con­crete (such as what the rule(s) are and how they were learned), which means such mod­els are less flex­ible… and thus more likely to tell us some­thing use­ful about the world, due to what they rule out.

  My ex­am­ples of sub­agents ap­pear­ing to mys­te­ri­ously an­swer ques­tions was meant to sug­gest that there are sub­tle things that IFS ex­plains/​pre­dicts, which aren’t au­to­mat­i­cally ex­plained in other models

  I don’t see that, though: if you ask peo­ple ques­tions, they are mys­te­ri­ously able to an­swer them, and the an­swers can of­ten be startling, and true. While I can’t speak to your sub­jec­tive ex­pe­rience, I can de­scribe some­thing from mine that sounds similar: if some­body asks me a ques­tion un­der cer­tain cir­cum­stances, I find my­self listen­ing to an an­swer com­ing out of my mouth that I did not know be­fore, and which I do not ex­pe­rience my­self as know­ing un­til af­ter I hear my­self say it.

  This state of af­fairs does not in­volve IFS-style parts, so ISTM that IFS does not add any­thing spe­cial here to the idea that ques­tions can trig­ger the ap­pear­ance of in­for­ma­tion in the mind that one did not ex­plic­itly know be­fore­hand… Al­most as if it had just been pre­pared fresh by a chef in the kitchen, vs. some­thing that was already in our re­friger­a­tor of knowl­edge. ;-)