# A memory is a physical thing. **URL:** \ **Category:** Great Debates\ **Created:** [March 2, 2005, 11:33am UTC](https://boards.straightdope.com/t/a-memory-is-a-physical-thing/292584 "2005-03-02T11:33:21Z")\ **Posts on this page:** 20\ **Page:** 11
**Author:** ![SentientMeat](https://avatars.discourse-cdn.com/v4/letter/s/43a26b/32.png) [@SentientMeat](https://boards.straightdope.com/u/SentientMeat)\ **Post date:** [March 5, 2005, 5:47pm UTC](https://boards.straightdope.com/t/a-memory-is-a-physical-thing/292584/201 "2005-03-05T17:47:56Z")
> [@o-w](#): > > if everything’s a continuum, we must be differentiating molecules and energy as such cognitively **We** do _everything_ cognitively, until we **die**. I’m afraid I _just don’t understand_ these responses of yours. > [@](#): > > Then maybe it isn’t so much whether memory is a physical thing; it’s whether cognition is a physical thing. As myself and others have pointed out, without interpretation, a memory isn’t a memory at all. Physicality is necessary for cognition, but is it sufficient? Do we _agree_ that a memory is a physical thing yet? If we **do** , would comparison of memories not be as physical a _process_ as death or cloud formation? > [@](#): > > Then the question becomes, what makes one neuronal link an association, while another is not? Because it literally, physically links the neurons storing those two memories (the visual memory and the auditory memory of the word). What makes a telephone call between two houses, rather than two _other_ houses? ---
**Author:** ![Liberal](https://avatars.discourse-cdn.com/v4/letter/l/848f3c/32.png) [@Liberal](https://boards.straightdope.com/u/Liberal)\ **Post date:** [March 5, 2005, 5:48pm UTC](https://boards.straightdope.com/t/a-memory-is-a-physical-thing/292584/202 "2005-03-05T17:48:59Z")
Incidentally, Digital, I do have a bookmark on Identity Theory, and I can share that with you: [http://plato.stanford.edu/entries/identity-personal/](http://plato.stanford.edu/entries/identity-personal/). Carter is mentioned briefly. ---
**Author:** ![Liberal](https://avatars.discourse-cdn.com/v4/letter/l/848f3c/32.png) [@Liberal](https://boards.straightdope.com/u/Liberal)\ **Post date:** [March 5, 2005, 6:09pm UTC](https://boards.straightdope.com/t/a-memory-is-a-physical-thing/292584/203 "2005-03-05T18:09:16Z")
> [@other-wise](#): > > .I’m curious; what do you make of claims of “non-dual” experience? That is, of consciousness without subject? I don’t know about any claims. Claims are just claims. But the assertion that Brahman cannot be the subject of Brahmaanubhava is deductively derived, and is valid. ---
**Author:** ![Digital\_Stimulus](https://avatars.discourse-cdn.com/v4/letter/d/aeb1de/32.png) [@Digital\_Stimulus](https://boards.straightdope.com/u/Digital_Stimulus)\ **Post date:** [March 5, 2005, 7:58pm UTC](https://boards.straightdope.com/t/a-memory-is-a-physical-thing/292584/204 "2005-03-05T19:58:12Z")
> [@Liberal](#): > > Try “W. R. Carter” + identity. For those interested in following up on this, I came across the following: A [list](http://www.imprint-academic.demon.co.uk/T2000/01-10.html) comprised of either classes or text summaries (I can’t tell which). Each is a pretty interesting little blurb in its own right; the one that mentions Carter is number 115. Also, there’s a [paper by Unger](http://www.nyu.edu/gsas/dept/philo/faculty/unger/papers/survival.pdf) (warning - PDF file) that delves into the biological and psychological theories of identity, including the dualism that possibly results. Personally, I have issues with philosophical essays of this type. I think my main complaint stems from the fact that words are fluid, their definition is often multi-faceted, and a subtle form of “moving the goalposts” occurs. This is perhaps unavoidable, as words are necessary to convey meaning (at least as concerns the medium of reading). At the same time, Unger has a legitimate point – akin to the point regarding memory gates changing states but not configuration. In my mind, it comes down to defining _continuity_. Take Unger’s example of “materially coincident entities” – the brass ball that he supplies with the names Barry and Patty (p21). Barry refers to the ball, while Patty refers to the brass. If we press the ball into wire, Barry disappears while Patty remains. If we progressively replace the brass with gold, Patty disappears while Barry remains. The continuity conditions for each is different, so this characterization is accurate. However, I think the problem is one of _definition_ – the fact that the meanings we assign are too poor to cover the gyrations we put them through (e.g., as **Liberal** said, “calling an amoeba aware is akin to stretching the fabric of an umbrella to cover a thousand heads when it was built to cover one”) is _our_ fault; one does not use a teaspoon to complete the “Big Dig” project, nor can one complain about receiving a nail instead of a screw when what was asked for was a “wall fastener”. There was a reference that I came across looking for Carter summaries (which I didn’t bookmark) that held that both biological and psychological theories of identity have something to offer and that by relying on both we can avoid the excesses of relying on either alone. That seems right to me, as an expansion of definitional term, so long as we’re aware when one part of the definition is no longer applicable. Is there a satisfactory answer? I don’t know, but I don’t think so; at least, not one that will satisfy people who are that deep into it. Is it a bad thing? I’d say not, as we do indeed glean something from pursuing these questions, if only an idea of where we fall short. Sorry for the mini-rant/hijack, if that’s what you consider it to be. ---
**Author:** ![Liberal](https://avatars.discourse-cdn.com/v4/letter/l/848f3c/32.png) [@Liberal](https://boards.straightdope.com/u/Liberal)\ **Post date:** [March 5, 2005, 8:11pm UTC](https://boards.straightdope.com/t/a-memory-is-a-physical-thing/292584/205 "2005-03-05T20:11:19Z")
> [@Digital](#): > > I think my main complaint stems from the fact that words are fluid, their definition is often multi-faceted, and a subtle form of “moving the goalposts” occurs. I can understand why you might think that, but I believe that we sometimes confuse equivocation with tedium. It has been my experience (and I have done much reading in several technical disciplines) that philosophers, on the whole, use words more precisely and with less equivocation than anyone else does. In fact, precision is a necessary tool for them. Unfortunately, sometimes comprehension suffers as a result of precision. Many of their writings are what Other-wise would call “dense”. It isn’t that they use words fluidly; it’s that they use them with as much efficiency as they can. Consider, for example, this sentence from Immanuel Kant: _Even the inner and sensible intuition of our mind (as object of consciousness) which is represented as being determined by the succession of different states in time, is not the self proper, as it exists in itself – that is, is not the transcendental subject – but only an appearance that has been given to the sensibility of this, to us unknown, being._ — Critique of Pure Reason Some philosophers (like Wittgenstein) have opted for more fluidity and therefore easier comprehension, but they are the minority. But as a price, the reader of Wittgenstein must fill-in-the-blank sometimes, and carefully consider exactly what was meant. ---
**Author:** ![Digital\_Stimulus](https://avatars.discourse-cdn.com/v4/letter/d/aeb1de/32.png) [@Digital\_Stimulus](https://boards.straightdope.com/u/Digital_Stimulus)\ **Post date:** [March 5, 2005, 8:12pm UTC](https://boards.straightdope.com/t/a-memory-is-a-physical-thing/292584/206 "2005-03-05T20:12:25Z")
> [@Liberal](#): > > Incidentally, Digital, I do have a bookmark on Identity Theory, and I can share that with you: [http://plato.stanford.edu/entries/identity-personal/](http://plato.stanford.edu/entries/identity-personal/). Carter is mentioned briefly. Thank you for that. It seems to me that to make any progress on whether memory is physical, we basically have to resolve the issue of identity. I’m intrigued by the “Simple View” mentioned in Section 3, as I feel the statement “a person here now is identical with a past or future being if and only they are identical” is exactly what I was getting at above with my point regarding definition. ---
**Author:** ![Liberal](https://avatars.discourse-cdn.com/v4/letter/l/848f3c/32.png) [@Liberal](https://boards.straightdope.com/u/Liberal)\ **Post date:** [March 5, 2005, 8:24pm UTC](https://boards.straightdope.com/t/a-memory-is-a-physical-thing/292584/207 "2005-03-05T20:24:16Z")
Just to follow up a bit on Digital’s point (while I wait for the steaks and biscuits to cook), probably the worst technical writing I’ve seen, in terms of fluidity and moving goalposts, is from biology and the life sciences in general. It is no wonder that a man of Other-wise’s intellectual acumen can fall so easily into the trap of anthropomophizing evolution. After all, biology articles and texts do it all the time — sometimes even in texts that are warning against it! I can’t recall the source, but I once encountered something like this: “Do not assign to evolution a guiding hand, for guidance implies an intelligence that nature lacks. Nature does what she does only because she can do no other.” 😃 Plus, it has always impressed me as a fine irony that the very thing biology studies — life — is just about the most equivocal term in all of science. Some texts pretend to define it, but no definition I’ve seen has ever been satisfactory to me. And although I’m no scientist, many scientific writers [agree](http://www.space.com/searchforlife/seti_shostak_020829.html). ---
**Author:** ![Digital\_Stimulus](https://avatars.discourse-cdn.com/v4/letter/d/aeb1de/32.png) [@Digital\_Stimulus](https://boards.straightdope.com/u/Digital_Stimulus)\ **Post date:** [March 5, 2005, 8:26pm UTC](https://boards.straightdope.com/t/a-memory-is-a-physical-thing/292584/208 "2005-03-05T20:26:56Z")
> [@Liberal](#): > > I can understand why you might think that, but I believe that we sometimes confuse equivocation with tedium. It has been my experience (and I have done much reading in several technical disciplines) that philosophers, on the whole, use words more precisely and with less equivocation than anyone else does. In fact, precision is a necessary tool for them. Unfortunately, sometimes comprehension suffers as a result of precision. Many of their writings are what Other-wise would call “dense”. It isn’t that they use words fluidly; it’s that they use them with as much efficiency as they can. Oh, og, don’t I know it. Heidegger was/is horrid. Worthwhile (in some respects) once you wade through it, but ugh, is it tough going. My point is not a criticism of philosophy, philosophers, or the texts; it is more a self-criticism. We converse like this with words under the assumption that the receiver will understand what we mean, that they share the same definitions we do. At best, however, definitions are only similar; at worst they don’t really coincide at all. As I said a couple times early on in this thread, I’m often accused of quibbling; however, I find that in most cases, it’s not quibbling so much as reaching an agreed upon definition. For matters of any depth, the price has to be paid in time and effort; the expense is often too much for the parties involved, and the bargain of half-price entry results in a dearth of actual communication. (My, how poetic I am today.) ---
**Author:** ![other-wise](https://avatars.discourse-cdn.com/v4/letter/o/4bbf92/32.png) [@other-wise](https://boards.straightdope.com/u/other-wise)\ **Post date:** [March 5, 2005, 9:04pm UTC](https://boards.straightdope.com/t/a-memory-is-a-physical-thing/292584/209 "2005-03-05T21:04:33Z")
> [@SentientMeat](#): > > Do we _agree_ that a memory is a physical thing yet? If we **do** , would comparison of memories not be as physical a _process_ as death or cloud formation? Nope. That’s why this thread is five pages long. 🙂 My point is that if memory is a form of information we cognize (roughly, **DS** ’s suggestion) and _if_ information is not strictly physical, than memory is not a physical thing. > [@](#): > > Because it literally, physically links the neurons storing those two memories (the visual memory and the auditory memory of the word). What makes a telephone call between two houses, rather than two _other_ houses? Let’s run with that analogy. You also could string Cat5 cable between two houses to link them. Cat5 wire and telephone wire can be physically distinguished from each other. How do you physically distinguish between a neuronal link that’s an association between two memories, and a neuronal link that is not? ---
**Author:** ![Liberal](https://avatars.discourse-cdn.com/v4/letter/l/848f3c/32.png) [@Liberal](https://boards.straightdope.com/u/Liberal)\ **Post date:** [March 5, 2005, 9:05pm UTC](https://boards.straightdope.com/t/a-memory-is-a-physical-thing/292584/210 "2005-03-05T21:05:24Z")
> [@Digital](#): > > Heidegger was/is horrid. Yes, Heidegger indeed. Hegel too. ---
**Author:** ![other-wise](https://avatars.discourse-cdn.com/v4/letter/o/4bbf92/32.png) [@other-wise](https://boards.straightdope.com/u/other-wise)\ **Post date:** [March 5, 2005, 9:19pm UTC](https://boards.straightdope.com/t/a-memory-is-a-physical-thing/292584/211 "2005-03-05T21:19:07Z")
> [@Liberal](#): > > I don’t know about any claims. Claims are just claims. But the assertion that Brahman cannot be the subject of Brahmaanubhava is deductively derived, and is valid. I didn’t think of that. I suppose that _would_ be self-contradictory (so to speak :)). Not being established in Brahmaanubhava is probably the only reason we can be the subject of ourselves. ---
**Author:** ![other-wise](https://avatars.discourse-cdn.com/v4/letter/o/4bbf92/32.png) [@other-wise](https://boards.straightdope.com/u/other-wise)\ **Post date:** [March 5, 2005, 9:28pm UTC](https://boards.straightdope.com/t/a-memory-is-a-physical-thing/292584/212 "2005-03-05T21:28:28Z")
> [@Liberal](#): > > Yes, Heidegger indeed. Hegel too. Someone once told me: "It’s impossible to translate Heidegger. Even into German.” ---
**Author:** ![Digital\_Stimulus](https://avatars.discourse-cdn.com/v4/letter/d/aeb1de/32.png) [@Digital\_Stimulus](https://boards.straightdope.com/u/Digital_Stimulus)\ **Post date:** [March 5, 2005, 9:44pm UTC](https://boards.straightdope.com/t/a-memory-is-a-physical-thing/292584/213 "2005-03-05T21:44:50Z")
> [@SentientMeat](#): > > Do we _agree_ that a memory is a physical thing yet? For the record, I believe that memory _necessarily_ has a physcial basis (or nature, as put forth in the OP). I don’t see the impediment to declaring this, unless one specifically defines “memory” as something that is cognized. Even then, I don’t see an issue with saying memory is physical; the issue is in the definition of awareness or cognition. I think **Liberal** is not convinced of this yet (again, I base this on [post #156](http://boards.straightdope.com/sdmb/showpost.php?p=5910325&postcount=156)), and I’m hoping some particular objection or at least means of furthering the discussion is forthcoming. ---
**Author:** ![other-wise](https://avatars.discourse-cdn.com/v4/letter/o/4bbf92/32.png) [@other-wise](https://boards.straightdope.com/u/other-wise)\ **Post date:** [March 5, 2005, 10:06pm UTC](https://boards.straightdope.com/t/a-memory-is-a-physical-thing/292584/214 "2005-03-05T22:06:19Z")
> [@Digital Stimulus](#): > > I don’t see the impediment to declaring this, unless one specifically defines “memory” as something that is cognized. **DS** , I’m confused. I can’t reconcile the above with this quote: > [@Digital Stimulus](#): > > Let me express it this way – a memory is a form of information that an aware being can interpret. ---
**Author:** ![Digital\_Stimulus](https://avatars.discourse-cdn.com/v4/letter/d/aeb1de/32.png) [@Digital\_Stimulus](https://boards.straightdope.com/u/Digital_Stimulus)\ **Post date:** [March 5, 2005, 10:55pm UTC](https://boards.straightdope.com/t/a-memory-is-a-physical-thing/292584/215 "2005-03-05T22:55:59Z")
> [@other-wise](#): > > **DS** , I’m confused. I can’t reconcile the above with this quote: There was more to that quote. Here it is in its entirety: > [@Digital Stimulus](#): > > I think you’re right in your response to other-wise. I would augment the response with reference to my statements above; the meaning of a symbol is arrived at by applying an interpretation to a physical configuration. > > Come to think of it, I may be using “information” in a too-strong sense (as the reference to Claude Shannon indicates), as I can see objecting via a chicken-and-egg argument (e.g., information is necessary for an interpretation, but an interpretation is necessary for information). Let me express it this way – a memory is a form of information that an aware being can interpret. A memory still obtains information status if it provides some impetus for change (e.g., a simple sensor/motor connection). One does not have to be aware of a memory for it to have potential information content. > > Did I include a contradiction in there somewhere? I’m not sure I worked that one out properly, nor that I expressed it well. Perhaps I should’ve left things as **Hoodoo Ulove** put them. As indicated by the ending, I’m not convinced I did a solid job getting my point across. You’re right to bring it up again. It is meaning or interpretation of a symbol (where “symbol” and “memory” are equated here) that may or may not be physical. That comes from requiring a cognitive being to assign the meaning or interpretation to a symbol. A particular configuration of physical memory elements that record something is a memory, whether or not it is given meaning or interpreted. Part of what characterizes these memory elements is that they _can_ change state while remaining in a particular configuration. This is how I’m resolving the continuity issue; whether it’s a good resolution or not is up for debate. The obvious examples of such memory elements would be RAM chips or neurons; if others present themselves, I’d like to hear about them. Yes, in some sense, this gives erosion of a rock the property of memory. Common usage, although imprecise and unacceptable for the work we wish it to do, says this is OK (e.g., “Memories of prehistory writ large in the landscape”). I would challenge anyone to describe a “memory” in a sensible fashion that does not have a physical basis. If someone can do that, I’ll gladly either retract my statement or change some of the definitions with which I’m working. 😃 At any rate, it seems to me (referring to the Carter references and the Unger article I cited before) that the problem with the “psychological identity” position derives from starting with a complex idea – that of “identity” – and attempting to decompose it into constituent parts, which is essentially what is happening when we discuss thought-experiments such as brain-in-a-vat. Just as in software design, the choice of how to decompose an application into constituent parts is somewhat arbitrary and can be done in an infinite number of ways. Perhaps I’m a reductionist at heart. However, like Dennett, I’m willing to grant the reality of “folk wisdom kinds” – so long as they are recognized as such and not misapplied. Does that make sense? Are there any problems with it that I don’t see? Does it clear things up? ---
**Author:** ![other-wise](https://avatars.discourse-cdn.com/v4/letter/o/4bbf92/32.png) [@other-wise](https://boards.straightdope.com/u/other-wise)\ **Post date:** [March 5, 2005, 11:25pm UTC](https://boards.straightdope.com/t/a-memory-is-a-physical-thing/292584/216 "2005-03-05T23:25:05Z")
> [@Digital Stimulus](#): > > There was more to that quote. Here it is in its entirety: > > As indicated by the ending, I’m not convinced I did a solid job getting my point across. You’re right to bring it up again. It is meaning or interpretation of a symbol (where “symbol” and “memory” are equated here) that may or may not be physical. That comes from requiring a cognitive being to assign the meaning or interpretation to a symbol. A particular configuration of physical memory elements that record something is a memory, whether or not it is given meaning or interpreted. Part of what characterizes these memory elements is that they _can_ change state while remaining in a particular configuration. This is how I’m resolving the continuity issue; whether it’s a good resolution or not is up for debate. The obvious examples of such memory elements would be RAM chips or neurons; if others present themselves, I’d like to hear about them. > > Yes, in some sense, this gives erosion of a rock the property of memory. Common usage, although imprecise and unacceptable for the work we wish it to do, says this is OK (e.g., “Memories of prehistory writ large in the landscape”). I would challenge anyone to describe a “memory” in a sensible fashion that does not have a physical basis. If someone can do that, I’ll gladly either retract my statement or change some of the definitions with which I’m working. 😃 > > At any rate, it seems to me (referring to the Carter references and the Unger article I cited before) that the problem with the “psychological identity” position derives from starting with a complex idea – that of “identity” – and attempting to decompose it into constituent parts, which is essentially what is happening when we discuss thought-experiments such as brain-in-a-vat. Just as in software design, the choice of how to decompose an application into constituent parts is somewhat arbitrary and can be done in an infinite number of ways. Perhaps I’m a reductionist at heart. However, like Dennett, I’m willing to grant the reality of “folk wisdom kinds” – so long as they are recognized as such and not misapplied. > > Does that make sense? Are there any problems with it that I don’t see? Does it clear things up? I agree with the second and third paragraphs – I see no problems at all. I _think_ I could say the same thing about the first paragraph, but a couple of questions, if I may. In the case of neurons acting as memory elements, what would you consider to be the state changes they undergo (while remaining in a particular configuration)? Would it be, for example, firing rates? ---
**Author:** ![Digital\_Stimulus](https://avatars.discourse-cdn.com/v4/letter/d/aeb1de/32.png) [@Digital\_Stimulus](https://boards.straightdope.com/u/Digital_Stimulus)\ **Post date:** [March 5, 2005, 11:41pm UTC](https://boards.straightdope.com/t/a-memory-is-a-physical-thing/292584/217 "2005-03-05T23:41:07Z")
> [@other-wise](#): > > In the case of neurons acting as memory elements, what would you consider to be the state changes they undergo (while remaining in a particular configuration)? Would it be, for example, firing rates? Because biological neurons are so complex and I’m so ignorant regarding their operation, I hesitate to be emphatic. At my level of understanding, I’d say the configuration is specified by neuronal connections, while state changes would be neuron activation (is that equivalent to firing _rate_?). ---
**Author:** ![other-wise](https://avatars.discourse-cdn.com/v4/letter/o/4bbf92/32.png) [@other-wise](https://boards.straightdope.com/u/other-wise)\ **Post date:** [March 6, 2005, 12:43am UTC](https://boards.straightdope.com/t/a-memory-is-a-physical-thing/292584/218 "2005-03-06T00:43:41Z")
> [@Digital Stimulus](#): > > Because biological neurons are so complex and I’m so ignorant regarding their operation, I hesitate to be emphatic. At my level of understanding, I’d say the configuration is specified by neuronal connections, while state changes would be neuron activation (is that equivalent to firing _rate_?). Kinda-sorta. Neurons are always active, and if the electrical activity of the neuron passes a threshold (usually determined by the type of signals it’s receiving from _other_ neurons), the neuron will fire. The rate at which the neuron fires will vary depending on the input. I don’t know if you’d consider a change in the pattern of electrical activity in a neuron or neuronal cluster to be the same as a change in it’s state. Would a change in the pattern of electrical activity in RAM chips be a change of their state? ---
**Author:** ![Voyager](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/voyager/32/133_2.png) [@Voyager](https://boards.straightdope.com/u/Voyager)\ **Post date:** [March 6, 2005, 12:51am UTC](https://boards.straightdope.com/t/a-memory-is-a-physical-thing/292584/219 "2005-03-06T00:51:08Z")
> [@Liberal](#): > > Voyager, I’m not sure I would call that “learning”. Learning has some rather daunting epistemic baggage. Perhaps “improving” would be a better term. It is commonly called learning. Improving does not provide enough information, since you can improve a program’s performance from outside, but this learning requires no change to the program as entered by anyone, and involves input from playing. There is no assumption of intelligence or consciousness, but then animals learn also, so the word should not have those connotations. Still, I think these programs do more along that line than you seemed to think they do. ---
**Author:** ![Voyager](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/voyager/32/133_2.png) [@Voyager](https://boards.straightdope.com/u/Voyager)\ **Post date:** [March 6, 2005, 1:07am UTC](https://boards.straightdope.com/t/a-memory-is-a-physical-thing/292584/220 "2005-03-06T01:07:23Z")
> [@Digital Stimulus](#): > > This points to the weakness of my reliance on the Ship of Theseus paradox. Even though measurements of the physical state are changing, the actual configuration of hardware isn’t changing (barring FPGAs, and evidenced by the fact that the results of dumping core will always exhibit the same structure). I’m not sure how else to handle the objection; any ideas? I have no answer to this - I think it depends on how you define identity. I actually just wrote something that is a practical application - if you identify a computer with a serial number, say, how do you handle repairs and upgrades to it? Do you give it a new one at some time, do you add serial numbers, or do you consider the computer with all new boards identical to the old one? I think it depends on what you want to do with the serial number. I even wrote about the case where everything is different but the chassis. My answer had to do with our goals, and would not be philosophically satisfying. During computations, you don’t change things (except for hot swapping, etc.) But in some sense the state of the hardware does change. There are various mechanisms that grow small defects intol bigger ones that might cause an error. There are all sorts of wearout mechanisms. No processor lasts forever. So, my answer to the paradox lies with General Semantics - there are two (many) ships with the same tag. > [@](#): > > An interesting point, I think. How might one characterize a “memory element”? What I mean is, I can see qualifying RAM and neurons as memory elements, but it is conceivable (I think) to have a memory element that is neither. What would it be like? Might a possible description be characterized by the idea expressed in the paragraph above – that there is some physical configuration that retains continuity but whose state can change without changing the overall configuration? Nifty…I like that one, as it supplies a concrete functionality that can be “tested” for. You pretty much have it - a memory element is one that holds a state (though it might require a refresh) while the power is maintained. Certain defects in memories cause state to be lost, and a retention test, which consists of loading a pattern, waiting for milliseconds (an eternity in my business) and reading the pattern is used to test for this. There is something called a bus keeper, but on a tristate bus to hold the value when no inputs are driving, which does not last very long, and which is not considered a memory element. FPGAs, by the way, have a memory plane - they are basically memory cells that control the configuration of the gates, but more statically than microcode or software and less statically than putting the control in pure hardware, like synthesis. [Previous page](https://boards.straightdope.com/t/a-memory-is-a-physical-thing/292584.md?page=10) [Next page](https://boards.straightdope.com/t/a-memory-is-a-physical-thing/292584.md?page=12)