Talk:Positive feedback/Archive 2

Hyperbolic world-system stuff
Does it make sense to have a section in this article on an obscure theory and application, sourced only to some guy's primary papers? I don't think so, which is why I've taken it out several times. Does anyone think it belongs? Dicklyon (talk) 05:44, 12 June 2012 (UTC)


 * You removed a section that had sources? Were the papers published and peer-reviewed? --Trevithj (talk) 06:16, 12 June 2012 (UTC)


 * Yes; it's only very tangential to the topic of the article, and seems to be part of a campaign of promotion of a fringe idea. Dicklyon (talk) 18:42, 12 June 2012 (UTC)


 * I have added a reference to Michael Kremer's "Population Growth and Technological Change: One Million B.C. to 1990", The Quarterly Journal of Economics (The MIT Press) 108 (3): 681–716, 1993, DOI:10.2307/2118405, JSTOR 2118405; is it OK now? Actually, Dicklyon, I would appreciate if you could read Kremer's paper before deleting this section again.
 * Phanerozoic (talk) 09:44, 12 June 2012 (UTC)


 * I don't see the paper accessible, but I read some summary/comments on it. The so-called historical tests of the hypothesis sound a lot like what Feynman called cargo-cult science.  Dicklyon (talk) 18:42, 12 June 2012 (UTC)


 * It might make sense to go to Malthusian growth model and introduce the concept of "hyperexponential" as an accelerating exponential growth, if that's a sourced meaning. Certainly hyperbolic growth isn't right, unless these nuts are predicting a singularity in which the Earth's population and carrying capacity go to infinity.  Dicklyon (talk) 19:45, 12 June 2012 (UTC)


 * The section title shows some bias, being capitalized and not being a term in common usage. The section could be moved under the "population and agriculture" section, without being out of place. I have done this. I have no other objections to the section than that though. I DO have objections to cited text being (re)worded in ways not supported by the citation. The original cited paper by Korotayev says "hyperbolic growth". --Trevithj (talk) 21:30, 12 June 2012 (UTC)


 * Yeah, that's the trouble with relying on flaky primary sources. Dicklyon (talk) 00:26, 13 June 2012 (UTC)


 * It doesn't appear to be a primary source - the original research dates back to the 60s, and the paper cites about 50 sources. Have you read this paper? If you have cited counter-examples to support your claims of "flaky" or "fringe idea" or "cargo-cult" or " mathematically silly model that's inconsistent with all potential realities", please add them to the section. Right now your changes have impacted the integrity of the cited source. --Trevithj (talk) 01:06, 13 June 2012 (UTC)
 * Two of the three cited sources are primary sources by Andrey Korotayev, and the third, by Kremer, is paywalled, so I haven't read it. The idea of hyperbolic growth is silly; a functional form that may fit in the short term, but whose logic consequence is illogical, an infinity in a finite time.  This is the idea that leads to the concept of quadratic positive feedback; if we want to explain that concept, that are probably better contexts for doing so.  On the other hand, I am unable to find any book hits for "quadratic positive feedback" and "hyperbolic growth" that are not written by Korotayev. Dicklyon (talk) 05:07, 13 June 2012 (UTC)
 * ?? Korotayev's 2005 paper begins with The fact that up to the 1960s world population had been characterized by a hyperbolic trend was discovered quite some time ago... and he cites 5 other sources dating back to 1960. He is hardly a primary source. However one of the 30 more recent articles that cite this paper does point out that [this model] does not mean population growth and innovation can accelerate indefinitely. If that is your point, I don't think anyone is arguing. The point is that the explanations offered invoke "positive feedback". --Trevithj (talk) 22:36, 13 June 2012 (UTC)
 * Having said that, I've modified the section to cite less primary sources that are more relevant to the article topic. --Trevithj (talk) 01:06, 14 June 2012 (UTC)
 * The problem I had was with the whole chain from an approximate trend to an inferred differential equation to an argument about causation in which higher population is the cause of more inventions, and the pretense of having tested that hypothesis. It just didn't seem that an article on positive feedback would benefit from such a strained example.  Dicklyon (talk) 01:20, 14 June 2012 (UTC)


 * OK, I fixed it to make it clear that the hyperbolic-like growth was only up to 1960, as Korotayev noted; I linked the cited paper/book. Dicklyon (talk) 05:28, 13 June 2012 (UTC)
 * Is 'apparently' the right word to use to describe what's in a source in Wikipedia's own voice? It's rather whithering isn't it? A bit like putting "Oh, really?" after a quote. Do we have any sources that support us rubbishing the maths in these papers? WP:SAID etc. --Nigelj (talk) 19:48, 13 June 2012 (UTC)
 * Whithering? If you don't like "apparently", some words from the source are "can be described" and "empirical estimates".  Nobody claimed that the hyperbolic was a possibly accurate model in the long run; that would be mathematically absurd.  Dicklyon (talk) 20:28, 13 June 2012 (UTC)
 * Isn't that the thing about positive feedback? In the 'soft' linear region it sets off heading absurdly for infinity, then real-world limits hit in (e.g. supply rails in electronics) and it stays finite in the end. I'm sure this paper wouldn't use the word (hyperbolic) without the author knowing what it means, and being able to justify it mathematically. --Nigelj (talk)
 * It is absolutely without justifying it mathematically. It is simply an empirical observation that the data sort of fit a hyperbolic model.  After, a quadratic feedback model was invented to give that result from a simple equation.  That's all OK, as far as it goes, but to say it in a way that makes it sound like a good model of reality would be going too far.  Dicklyon (talk) 03:15, 15 June 2012 (UTC)

US cost of living adjustment
I just reverted some edits by. If reliable commentators are relating this to positive feedback and its effects, then it will make an excellent addition to the article. However, we need a balanced report on what leading thinkers are saying, with cited sources. We also must get by without saying things like "all financial plans that incorporate it are doomed" as if that were Wikipedia's own considered opinion on the matter - Wikipedia never expounds such opinions, but may quote others who do. --Nigelj (talk) 23:33, 5 December 2012 (UTC)

Amplification
Amplification is a well defined term. It refers to using a small signal directly to increase and decrease a larger signal from a separate power source. In electronics, for example, an increase in the small current flowing into the base of a transistor (e.g. from a microphone) results in an increase in the much larger current being drawn from a separate power supply into the transistor's collector. If this larger current is made to flow through a loudspeaker, then the sounds picked up by the microphone may be getting amplified.

Positive feedback is also well defined, as described this article, for example 'A produces more of B which in turn produces more of A'. In amplification there is the central concept of gain which is by definition more or less constant. In positive feedback we often see exponential growth, which would imply exponentially growing gain, dependent on input level, which is absurd in the context of any practical amplifier.

Twice recently I have reverted a parenthesised addition to the very first words of the WP:LEDE that seems to say that positive feedback is a kind of amplification. This is a muddled use of defined terms, and I'm afraid that no scholarly, reliable, technically sound reference will ever be found to back it up. --Nigelj (talk) 14:15, 31 December 2012 (UTC)

Thanks Nigelj. You're likely correct on no sound reference existing.

However, every single time I try to explain Positive and Negative feedback to any layperson, they immediately begin by thinking "positive = good" and "negative = bad".

Yes, I realize that positive feedback is a well-defined term...but I think it is un-necessary for the entire academic world to always re-explain that positive feedback is not necessarily "good".

Amplification is not a well-defined term in my opinion. There's not even a unique wiki-page on it...but actually multiple wiki-pages(http://en.wikipedia.org/wiki/Amplification).

When I personally explain positive feedback loops in economic theory, I use the term amplifying feedback loops instead of positive feedback, so that I don't have people being immediately confused.

I don't think that amplification needs to be unique to electronic signaling.

Thoughts? I really believe this small parenthesized change is worthwhile to the academic discussion of systems theory.

EzekielChang (talk) 20:30, 31 December 2012 (UTC)
 * Are you the same user as Sapientsage and 76.167.33.232? If so, I think you should have a look at WP:SOCK. WP:V answers your main point - "Wikipedia does not publish original research" - and what you say about what you find personally counts as "original research" unless it has been published by a reliable academic in the subject area. --Nigelj (talk) 20:51, 31 December 2012 (UTC)

Hi Nigelj --

Yes, I am Sapientsage, and I wasn't trying to create multiple accounts to avoid detection. The IP edit was because I forgot to log in initially when I changed one, but then remembered to log on before changing the other.

I figured it would be pretty obvious it was the same person given that I tried to make corrections based on the reasons you gave for removing my updates.

My Sapientsage account was deactivated by another Wiki editor (Alexf) because the user name is too similar to my website. This is why I switched to my real name. I have kept my same email address. Please note that I have not attempted to re-make the edits that you removed under my Sapientsage username.

I am new to posting on Wikipedia.

The rules make sense; so hope you don't mind as I am still learning what all the rules are.

Thanks,EzekielChang (talk) 00:54, 1 January 2013 (UTC)
 * Hi Ezekial. There's nothing wrong with being new - there's a lot to learn around here! Have a look at WP:SOCK and add the relevant links on your two user pages to avoid any confusion or doubt as soon as possible. As to the rest, do have a read of WP:V - it is pretty central to what we do here. I left a link to the 'five pillars' on your previous user talk page as part of a standard 'welcome' message. If you can find reliable sources that make points that you find useful, they can be added to the articles. Maybe initially to the body of the article (below the table of Contents). What goes into the WP:LEDE is pretty important - and the first sentence is almost sacrosanct(!) - there's a whole section on it at that link. Finally, if what you add is reverted or removed, don't just re-add it, but always open a discussion on the relevant talk page like we have here, to avoid any hint of an WP:EDITWAR. Happy New Year, and happy editing. --Nigelj (talk) 01:19, 1 January 2013 (UTC)

Got it. Thanks for all the information. I'll go through it before making my next edit. Thanks Nigelj! EzekielChang (talk) 05:03, 1 January 2013 (UTC)

Parameters and rates of production
I'm not happy with all the changes that were made in this edit. First of all it introduced and made significant use of the word parameter. To me, the parameters of a system are variables that can be changed, and which when changed, affect the way the system works. They are like settings or adjustments. The gain of an amplifier, the derivative feedback timeconstant, these are the parameters within which the system has been set to operate. Therefore "a change in the value of a system parameter ultimately causes an increase in the magnitude of that change" makes me think that the settings are changing, rather than the process and manipulated variables.

Secondly at a crucial point in the definition it introduced the phrase "the rates of production of "A" and "B"". This is suddenly to introduce the first derivative of the variables under discussion, where there is no need to do so.

I think people have trouble understanding that these definitions and discussions in easy everyday language are actually technically precise in a precisely defined mathematical world. The fact that they still read like ordinary English is testament to the amount of effort that has gone in, over the years here, to refining and simplifying the language used without compromising the technical and mathematical accuracy of what is actually said. With my tongue firmly in my cheek, sometimes I wonder if people think that it sounds too straightforward, and thinking like a scriptwriter from Dr Who or Star Trek, wonder if it might be more impressive if it had a few references to 'amplification', 'parameters', and 'flux capacitors'.

There were some grammatical improvements in the edit in question, and so I'm going to try to partially revert it, keeping the best bits. If would like to show us references that support the wording he added, I'd be glad to see them. Glad and very surprised. --Nigelj (talk) 10:56, 4 February 2013 (UTC)
 * Sorry. Unfortunately, even if I weadled out the word parameter, I still couldn't keep the phrase "a change in the value of a system parameter ultimately causes a decrease in the magnitude of that disturbance." The problem was the implication of time delay and the use of word ultimately. Feedback often happens instantaneously (well, at the speed of light). It is usually the system upon which the feedback acts that has the sluggish 3rd, 4th, or higher order, delayed responses. It's another one of these tricky points that it's better not going into in an opening paragraph. The original statement that they "act to reduce or counteract it" carefully does not imply one way or the other whether the feedback action is instantaneous or whether it won't have its full effect for hundreds of years. --Nigelj (talk) 11:11, 4 February 2013 (UTC)

Compressive and expansive amplifiers
For electronic circuits there are two cases: The amplifier has decreasing gain with increasing input amplitude and the far more common case where the amplifier has increasing gain with increasing input amplitude (until the amplifier must limit due to overload). The first case allow smooth transition into and out of oscillation (if part of an oscillator circuit) and the amplitude of oscillation is somewhat controllable by adjusting the gain of the amplifier. In the second case the circuit will bust into oscillation at some point just below where the loop gain is 1 (kicked above 1 by noise) and only limit by amplifier overload, the amplifier gain having to be reduced substantially to stop oscillation. — Preceding unsigned comment added by 113.190.162.57 (talk) 12:38, 15 April 2013 (UTC)

The derivative only?
Under 'Overview', the article says, "In positive feedback, the derivative of the variable is positively affected by the variable's value, and the opposite is true in negative feedback." I don't think that this is a good description of all positive feedback systems. I do not have a copy of the book referenced, but I find it unlikely that positive feedback has to be restricted to cases where only the first derivative is affected by feedback. The error seems first to arise in the previous para which says that a feedback system is "a system [that] has a feedback loop influencing its own rate of change." I wonder if someone who does have access to this book would like to comment here, before we rewrite these descriptions to be more general and more true. --Nigelj (talk) 19:30, 15 April 2014 (UTC)


 * Does the book link not work for you? That book focuses on feedback of error to derivative, which is pretty conventional; it doesn't say "only", and the dynamics are of course usually more complicated than just this first-order description.  Dicklyon (talk) 19:40, 15 April 2014 (UTC)
 * The problem is that in the online preview, there are so many 'unavailable' pages that it's difficult to see the point of the examples being created in Chapter 3. There is no doubt that the chapter is concerned with modelling systems involving the derivative of a variable being fed back. What is not clear is why the authors are considering this here. To define all feedback systems - positive and negative - in this way is clearly in contradiction to the lede of this article ("Mathematically, positive feedback is defined as a positive loop gain around a feedback loop.") and with the diagrams adjacent to the statements. In the simple stampede example, it does not say that the "Overall level of panic" affects the "Rate of increase in the number of cattle running". In the Schmitt trigger circuit, there is a simple resistor feeding the output back to the input, not a capacitative differentiator. The same in the R-S flipflop - no capacitors in the feedback. And so on. The cited book is quite correct when it says, "A positive feedback loop is one with an even number of negative influences [around the loop]," as quoted in the ref, but it is strangely unhelpful in its insistence on derivative actions in Chapter 3's modelling examples, IMHO. --Nigelj (talk) 20:26, 15 April 2014 (UTC)
 * I don't have the book. The lead definition is also too simple to be correct in general, since loop gain is actually complex and frequency dependent.  Putting in the derivative suggests at least the direction of phase shift that's relevant.  Dicklyon (talk) 20:44, 15 April 2014 (UTC)
 * Complex quantities can still be 'positive' on the real axis. Frequency dependant, yes, but if it's positive at this frequency, then there is positive feedback at this frequency (it can be temperature dependent too, but the same applies). "the derivative suggests at least the direction of phase shift"? No, you've lost me there. I have no idea what that might mean in this context. It appears that you are changing the subject, and quite frankly I'm not interested in a long, rambling debate. Are you telling me that this is the only definition of positive feedback that you think is correct? And that you want to change the lede to include this too? Are you prepared to engage with the examples I gave above where positive feedback exists without a derivative? --Nigelj (talk) 21:34, 15 April 2014 (UTC)
 * To try to focus this discussion onto something concrete, I have made an edit to show the kind of rewrite that I have in mind. I have combined several single-sentence paragraphs into one; I have kept much of the original text, but removed the requirement of the first derivative being used in the feedback as discussed above. I have added a sentence about value judgements, as I know that this is a common mistake made by people who first hear the term in, e.g. a business or educational context. I have added two references, but kept some text referenced to the book discussed above. --Nigelj (talk) 22:25, 15 April 2014 (UTC)

Unrelated topics
I have deleted three tiny sub-sections from the article recently, as, as far as I can see, they do not describe a mechanism in which a positive feedback loop, as described in this article, can be identified in the description or in the reliable sources. Maybe we should require that sources used to introduce new topics at least mention the term 'positive feedback' somewhere in the text. --Nigelj (talk) 20:46, 4 May 2015 (UTC)
 * Human sense of sight: This unreferenced section seemed to try to draw a connection between reading damaging your eyes, reading increasing your IQ, and an increased IQ making you want to read more, which damaged your eyes even more than before. I don't think any of this can be supported by ophthalmic, academic, medical or educational research, and there is certainly no one I can find calling this (non-existent) effect a positive feedback loop.
 * Substance dependence and addiction: This section fairly accurately described the mechanism of acclimation among drug addicts, and the devastating effects this can lead to, but none of the Wikipedia articles links mentioned positive feedback. The addict's body acclimates to the drug, they take more, overdose and die. But for positive feedback to occur, the acclimation would have to force them to take more, when actually they choose to take more. The addictive personality and possible suicidal tendencies have important parts to play. Who is to say that the addictive personality is created or strengthened by the increasing doses? If the medical world find this a helpful way to explain drug overdoses to at-risk patients, or to their grieving relatives, then it will be easy to find the references to those recommendations. If not then I don't see how our unreferenced WP:OR will help anyone, especially after considering WP:MEDRS.
 * Internet recommender systems: This section seemed simply to muddle the idea of people writing or ticking that they liked a product they had bought (i.e. giving 'feedback' that is 'positive'), with the actual concept of a positive feedback loop. Giving a product marketing website 'positive feedback' about the products they sell may lead somehow to you buying more stuff, but I think the confusion of the two terms is too prominent to make anything encyclopedic about this. If this is established knowledge about recommender systems, then let's just see a few quotes from the published sources where the term 'positive feedback' is explicitly mentioned.

Assessment comment
Substituted at 03:16, 30 April 2016 (UTC)

Suggest image for intro
You might want to change the caption.--Guy vandegrift (talk) 18:30, 10 November 2016 (UTC)
 * This is not better than any of the current three images in the lead section, in my opinion, if that is the question. It is too mathematical; audio PA howling is not the best introduction to the wider concept; it might be possible to work this in further down the article in the sections where audio and video feedback are discussed. --Nigelj (talk) 23:30, 12 November 2016 (UTC)
 * This is a class project with 28 Wikiversity newbies-we will probably insert this down in the audio/video section in a class discussion.--Guy vandegrift (talk) 17:25, 15 November 2016 (UTC)
 * We decided that this image was too complicated for the lede in this article, but too simplistic for the "audio section".  This image is too simplistic b/c it doesn't describe that single-frequency pitch one hears.  We chose this model for its algebraic simplicity -- perhaps to pedagogical for wikipedia?  When we make a nice Wikiversity article, we might try to link out of it with a sister link. --- matsief1, jason sheer, cantelli_25, Boss429, balllikeimkobe, w062bls (we are six of the 28 Wikiversity newbies) Authored with help by -Guy vandegrift (talk) 18:44, 30 November 2016 (UTC)

The feedback loop or depression
Did I miss it? Or was it not included? I think it would fit well in this page. The concept as best put by Mike Myers. "I can't stop eating. I eat because I'm unhappy, and I'm unhappy because I eat.  It's a vicious cycle" --172.114.196.164 (talk) 05:02, 1 June 2017 (UTC)

External links modified
Hello fellow Wikipedians,

I have just modified one external link on Positive feedback. Please take a moment to review my edit. If you have any questions, or need the bot to ignore the links, or the page altogether, please visit this simple FaQ for additional information. I made the following changes:
 * Added archive https://web.archive.org/web/20131008160618/http://www.sustainabilityinstitute.org/pubs/Leverage_Points.pdf to http://www.sustainabilityinstitute.org/pubs/Leverage_Points.pdf

When you have finished reviewing my changes, you may follow the instructions on the template below to fix any issues with the URLs.

Cheers.— InternetArchiveBot  (Report bug) 18:26, 4 June 2017 (UTC)

Conservation
I added a section briefly explaining how endangered wildlife can become extinct by human actions which becomes a positive feedback loop. I probably didn't word it very well but I added a reference. If someone wants to rework it in a better manner feel free to! Xanikk999 (talk)

Only available through the Roblox Chromebook App. — Preceding unsigned comment added by 125.161.113.181 (talk) 07:27, 28 February 2019 (UTC)

Positive feedback
I added /ref> Such positive feedback loop can lead to an economic bubble.Komlos, J. Foundations of Real-World Economics. page 74-76. Routledge 2019. what do you have against this addition? I linked to economic bubbles and i added a reference where this is discussed in more detail. please let me know what you find objectionable about it. i am conveying an important piece of scientific information. — Preceding unsigned comment added by Jkomlos (talk • contribs) 19:52, 5 April 2019 (UTC)

Hysteresis
The section on Hysteresis needs some work, especially the last part on bistable behaviour. It's not correct to say the sign of the feedback reverses. But how should this be described? SaffyLane (talk) 03:33, 14 July 2021 (UTC)