What Is A Tulpa – In Detail

(This is an essay that was jointly written by a contributor, Falah, and their tulpa Noctis and originally posted elsewhere. All views expressed are the opinion of the authors.)

The Basics: Definition and History

So, let’s start with the metaphorical dictionary. In the broadest use of the word as is found here:

A tulpa is an autonomous entity existing within the brain of a “host”. They are distinct from the host in that they possess their own personality, opinions, and actions, which are independent of the host’s. A fully-formed tulpa is, or highly resembles to an indistinguishable point, an actual other sentient, sapient being coinhabiting with the host consciousness.

Tulpamancy originally came from an esoteric Tibetian Buddhist practice that was discovered by internet-goers. However, the practices the tulpamancy community (hereon referred to as “Western tulpamancy”) devised, and its beliefs, are extremely different from Tibetian tulpamancy. The ends to which tulpas are created, beliefs regarding their nature, relationship between tulpa and tulpamancer, attitudes, approaches, and methodology are all different. Without going into much detail, it is safe to say that for the most part, the only connection Western tulpamancy has to Tibetian tulpamancy is the usage of the name and original inspiration.

What Tulpamancy Is Not

It’s not how it’s described in the creepypastas. Very, very few tulpas turn “evil”, much less try to harm their hosts. There are certainly disagreements and even misunderstandings between tulpa and tulpamancer, just as there are between physical individuals, but that’s a far cry from the creatures in the creepypastas who just seem to be driven to murder on principles.

It’s not something that creates another physical individual. All metaphysics aside, tulpamancy is a phenomenon of the brain. The only way a tulpa can manipulate physicality is if they use possession or switching to control the host’s body, or if the host proxies for them.

It’s not DID (Dissociative Identity Disorder) or OSDD (Other Specified Dissociative Disorder). Tulpamancy involves sharing one’s head with others, and body control can be switched between those others, but that’s about where the similarities end. DID/OSDD, or clinical multiplicity, is a world of difference away from tulpamancy when it comes to culture, origins, and experience. DID/OSDD are disorders that cause significant distress, dysfunction, or danger. They overwhelmingly originate from childhood abuse and trauma, and thus those with DID/OSDD also face lasting, even debilitating struggles with PTSD, depression, and other co-morbid disorders. The DID/OSDD community is largely one of survival and recovery, in contrast to the tulpamancy community. It is very important not to conflate the two.

It’s not schizophrenia. Schizophrenia is a serious mental disorder consisting of breakdowns in the brain’s ability to process physical reality. Tulpamancers are perfectly aware that their tulpas are nonphysical, even when imposed. Schizophrenia doesn’t even always involve hearing voices, and studies have shown the disease to have biological and genetic bases.

It’s not any sort of “self-inflicted mental illness”. Putting aside the rather outlandish notion that mental illnesses can be “self-inflicted” to begin with, I’ll quote the DSM-V, the “bible of psychiatry” directly on this one:

A mental disorder is a syndrome characterized by clinically significant disturbance in an individual’s cognition, emotion regulation, or behavior that reflects a dysfunction in the psychological, biological, or developmental processes underlying mental functioning. Mental disorders are usually associated with significant distress in social, occupational, or other important activities. An expectable or culturally approved response to a common stressor or loss, such as the death of a loved one, is not a mental disorder. Socially deviant behavior (e.g., political, religious, or sexual) and conflicts that are primarily between the individual and society are not mental disorders unless the deviance or conflict results from a dysfunction in the individual, as described above.

In fact, many tulpamancers report an increased quality of life after creating their tulpas!

With the exception of its original context, it’s not a religious practice unless one makes it so. Tulpamancers come from all walks of life and a wide array of faiths. There are many ways tulpamancy and religion can interact, and a number of interesting occurrences in religion that involve hearing voices or being possessed by “external” forces, but a full discussion of this is not the purpose of this post.

It’s not a panacea for mental illness and personal problems. Like physical individuals, tulpas cannot and should not be expected to rid you of depression, anxiety, etc. They can make recovery easier, as voices of support and metaphorical shoulders to cry on, but they cannot outright cure it. Please see a professional if you are struggling from illness.

It’s not a way to access factual or experiential information that hasn’t been physically observed. Again, all metaphysics aside, tulpas are not separate physical beings. They cannot fly down to the library and transmit the contents of an encyclopedia to you, or tell you what your neighbor down the street is doing. In addition, creating a tulpa of a certain identity is not an accurate way to understand that identity. To quote what I told another curious member:

Some people say they’ve made or want to make a tulpa of another gender in order to understand how that gender thinks. I disagree quite strongly with this line of reasoning. A tulpa, all metaphysics aside, can only know what’s already in the brain. Many of the gaps in experience between genders come from the different ways in which each gender is raised, conditioned, and treated by culture. Even if your tulpa is of a different gender than you, they will not fully understand what it is like to live as that gender in the physical world unless you have experienced it yourself (e.g. being a trans person who has undergone social transition). Hell, even no individual member of a gender can fully understand what it is like to live as X gender in the physical world, because of how diverse culture is across different contexts. If you really want to understand another gender’s experiences, don’t make a tulpa. Read into the academia–it’s not just a “tumblr thing”, it’s a serious field of anthropology that goes back years–and talk to physical people of that gender, from many different backgrounds.

That being said, do not underestimate the power of permutation. Remarkable insights can be gained simply from remixing existing ideas into surprising combinations. As tulpas have different personalities from their hosts, they can look at the same concept from different angles and provide valuable insight in that manner, especially as time goes on and they have the opportunity to accumulate more of their own experiences, thus diverging further.

It’s not a way to become best friends with “[fictional character] in real life”. It’s perfectly okay for tulpas to adopt the names, forms, and even personalities of fictional characters, but characters from a tulpamancer’s own writing aside, they will not literally be that character from that world. Attempting to force a tulpa into being someone they aren’t will only end miserably for both tulpa and tulpamancer.

It’s not a way to create the boyfriend/girlfriend/significant other/sex slave of your dreams. I really hope this is self-explanatory.

It’s not a way to create a “perfect” friend. Tulpas run on the same logic as physical individuals, which means that they will not be perfect (if perfection even was a concrete reality instead of a wholly arbitrary and personal standard). Though you have tools to resolve them that you do not have with physical friends, there will be disagreements despite sharing a head.

It’s not a way to get special acclaim. So you made another person in your head? Great. Now go make a bridge, a new score of music, a painting, or an organization to solve any of the numerous ills blighting the world. You know those obnoxious soccer moms who can’t seem to shut up about how they deserve extra special reverence because they procreated? Don’t be the tulpamancy equivalent of that.

It’s not a way to replace other forms of social interaction, or to escape from one’s responsibilities. As discussed earlier, there are insights a tulpa cannot provide you with. If your goal is personal growth or learning, interaction with people who do not share a head with you is invaluable. Not to mention that tulpas, unlike physical friends, cannot drop by the store for aspirin when you’re sick, or help you move futons up the stairs. As for tulpas not being a way to escape from physical responsibilities… hopefully that goes without saying.

How Are Tulpas Made?

There are two aspects to this question: methodology, the specific methods by which a tulpa emerges, and mechanism, the underlying processes–abstract and physical alike–that are responsible for this formation.

First, methodology. Forcing a tulpa can essentially be boiled down into the process described in this forum post by Keiretsu:

Step 1. Read all the guides. Every one you can find.

Step 2. Throw out all the guides. Those are how others did things, not you.

Step 3. Think to and at your tulpa. Experiment with different methods learned in 1 and 2 until you find what you’re comfortable with. Repeat until your brain catches on and starts replying.

Step 4. Enjoy your tulpa.

That’s it. The whole step by step guide.

It wasn’t hard. Tulpa creation isn’t. The only hard part is that ‘repeat’ part. Seriously, it’s going to require repetition. How much repetition is going to depend on your own wiring, repetition duration and frequency, methods that work for you, and so on.

Don’t take anyone else’s step by step guide and hold it as the bible of creation. This is a deeply personal process, and each guide was a single person’s process. You aren’t them. There might be worthwhile pointers to use, but in the end, it’s going to be your own process that you develop.

Essentially, tulpa creation is a highly, highly subjective process. There is no set step by step process, no set timeframes for what milestones will happen when in what form. The first communication can be a headpressure, a word, several words, a slew of strange emotions, or something else entirely. You may receive this first communication after two weeks, a month, or several months. You may find visualization difficult, and find it beneficial to jump right into narration from the start. You might find narration difficult without an idea of the person you’re talking to, and need to construct a personality first. Or you might not. You might even have to dip into controversial methods like parroting, or being strict with what you accept as a response if your mind is naturally overactive and full of stray intrusive thoughts.

The key, above all, is persistence. Persistence, building up the presence of your tulpa being there, minimizing unproductive doubt–in most cases, if you had to ask whether it was you or not, it wasn’t you. Keep treating your tulpa like they’re there–albeit perhaps sleepy and “waking up” gradually–and eventually, they will be there.

Once a tulpa begins communicating consistently and coherently, your job becomes much easier. From that point on, it’s less a creation process from the host and more a joint effort between host and tulpa. Conversations become much easier when the other party holds their end–don’t worry if it’s rough at first, practice makes perfect. It’s also recommended that young tulpas have the opportunity to build up their own experiences by talking to others outside the body. Beyond that, it’s up to each group to decide whether they would like to pursue advanced techniques such as imposition, possession, and switching.

Now, for the thornier question. Mechanism. I will discuss this topic from a purely psychological view, as metaphysics is not my area of specialty.

The blunt truth: we have only scratched the metaphorical surface of neuroscience. We are further from understanding the inner workings of the brain than we are from understanding the workings of Mars. The brain is an infinitely intricate construct, far more complex than one area activating in response to one stimuli to do one specific task. In addition to the density of neurons themselves, we must consider also how those neurons are connected, in what combination, what neurotransmitters they send, and many other factors. We must also consider how areas of the brain interact in combination, what brainwave frequencies they generate, the degree of myelination of neurons in different areas, the proportions of ambient neurotransmitters in the intercellular medium… The brain is so incredibly complex that researchers are still discovering new networks and new methods of communication between areas of the brain.

Thus, no one has all of the answers here. Not neuroscientists, not psychologists, and most certainly not us. We are all looking at a collection of incredibly varied, incredibly intricate black boxes, and guessing at ways in which all of them might be related.

That being said, my personal, highly abstracted model is that tulpamancy is fundamentally a form of dissociation. To put it overly simply: a host dissociates from processes in their brain through forcing methods, and eventually those dissociated processes clump together, hit a certain critical mass, and up and become a person. Intrusive thoughts are stray, dissociated processes without a critical mass to cling to–they do not possess coherency, consistency, or identity, all hallmarks of a tulpa. Tulpas themselves are clusters of dissociated processes that have associated with each other to the point that they become self-referencing partitions of memory, which become more and more self-referencing as time goes on and they accumulate their own experiences. And hosts themselves are no different–hosts are their own cluster of processes and memories associated with each other, simply another self-referencing partition in the same brain with a unique sense of self, albeit the one that was there first.

There’s a mistake that people often make when they hear “dissociation”, which is to assume it simply means “someone fooling themself into thinking it’s not them.” It’s nowhere near that simple, as brains are nowhere near that simple. There are studies on DID systemsshowing that different members of a DID system can have, among other things, different allergies, different degrees of eyesight (to the extent that some may even be completely blind while others are sighted), and speak different languages. All show the brain’s very capable of partitioning to an extent that goes beyond “just fooling yourself”.

Granted, that’s DID, and as I’ve said, it’s important not to confuse it with tulpamancy. That being said, though, now that we’ve seen that brains possess the capability, is it then possible for tulpamancers and tulpas to partition that much? I say it is, from what I’ve seen in older tulpamancy systems and what I’ve seen myself. It simply takes longer without trauma to kickstart it, and given that a lot of groups seem uninterested in significant body timesharing, that degree of partitioning might never happen for some. But in my opinion, it doesn’t make them “less real” than for the ones who it does happen to. (In which I use “real” to mean “it exists, even if nonphysically, and has an impact on the world”.)

In any case, there is a great amount of speculation on the mechanisms of tulpamancy, and no shortage of different ideas. However, I do believe that these ideas may not be as incompatible as they may seem. Without going into too much detail, I view the matter as akin to the three blind men and the elephant, with the added complication of each blind man using different terminology to refer to the elephant.

At the very least, the matter is far from a simple one. We may never know in our lifetime, or in the next generation’s lifetime, the exact neurological processes that give rise to tulpas.

The Million-Dollar Question

So we’ve addressed methodology and mechanism. At first glance, it might seem that that’s all there is to understanding tulpas.

However, neither methodology nor mechanism really answer one of the oldest questions in the tulpamancy community: are tulpas actually conscious?

Before we can address this question, we have to address the question of what “consciousness” itself is. Even in science, it’s a complicated and often confusing word, whose meaning changes depending on what discipline you’re in. Some use it to refer to a hypothetical set of physical processes in the brain itself, and debate which mechanisms are necessary for the generation of awareness. Some use it to refer to the subjective, abstract experience of being, and raise questions concerning the nature of selfhood. Some use it to refer to a set of externally exhibited qualities, and ask what set of qualities an entity must exhibit in order to be considered conscious. The one common line through all of this–we don’t know enough to give definite answers, and abstract experience is something different from the physical mechanisms that give rise to it.

For the sake of this discussion, I’ll break it down and use a few different terms to cover this area. I’ll define the terms I use this way:

Now, there are many beliefs concerning tulpas and consciousness. I will describe three in brief.

The first category of belief holds that tulpas do not possess consciousness. They are simply tricks of perception on behalf of the host, in which the host associates various processes of theirs as belonging to “someone else”, until this self-trickery becomes automatic.

The second category of belief holds that tulpas are not tricks of perception, but rather parts of the host’s unconscious mind drawn into conscious awareness and personified. They are more than simply delusions, but in the end, they are all “part of the host”.

The third category of belief, and the belief held most widely in the community, is that tulpas are their own consciousnesses, possessing awarenesses distinct from the host’s. Rather than being parts of the host’s unconscious, they are consciousnesses of their own by virtue of their ability to process complexity and subjectivity in ways distinct from the host.

Which model is “correct”? No one knows. Awareness especially is impossible to observe aside from your own, hence the philosophical concept of the p-zombie, debates over whether human-like AI would experience awareness, and the perpetually unsolved “hard question of consciousness”, of how physical matter can create the abstract, immaterial experience of being. Until the hard question is solved and a definite way to physically measure immaterial experience can be devised–something unlikely to happen within our lifetimes–then we can neither prove nor disprove awareness in tulpas, no more than a host can prove that they are aware and not a p-zombie to another individual.

As for my personal thoughts on these models… I have put a great deal of thought into this subject, and there is much on it I would like to divulge. However, space is limited, and much of it would go beyond the scope of what is meant to simply be an overview. Thus, regrettably but necessarily, much of this will be cursory and greatly simplified.

In my opinion, the model that tulpas do not possess any form of consciousness or awareness, and are “only hallucinations”, is flawed in that it vastly oversimplifies the brain. In addition to what I have discussed earlier regarding the complexity of dissociation beyond “just fooling yourself”, the category works off an unproven assumption that the brain is singular and may only ever house one consciousness, implying that consciousness itself is one solid chunk, that the totality of the brain must be devoted to the generation of this consciousness, and/or that there is a singular core “consciousness center” in the brain. In fact, research finds the brain to be highly modular, more multiple than singular. Some of the most striking examples are split-brain patients, whose corpus callosums (the part connecting the hemispheres of the brain) are severed, in some instances physically separating the halves entirely. Each severed half continues to possess separate perceptions and actions, even to the point that they would be fighting over control of the body:

Gazzaniga and Sperry’s split-brain research is now legendary. One of their child participants, Paul S, had a fully functional language center in both hemispheres. This allowed the researchers to question each side of the brain. When they asked the right side what their patient wanted to be when he grew up, he replied “an automobile racer.” When they posed the same question to the left, however, he responded “a draftsman.” Another patient pulled down his pants with the left hand and back up with the right in a continuing struggle. On a different occasion, this same patient’s left hand made an attempt to strike the unsuspecting wife as the right hand grabbed the villainous limp to stop it.

As illustrated by this example, consciousness in action and sapience is an extremely complicated matter, far beyond singular “cores”–to quote a neuroscience friend, the brain is not so much one as it is many aligned to create the illusion that they are one, even to themselves. If sapience and mechanism in consciousness are far from simple, then I believe awareness, the abstract product of these processes, is likewise far from simple.

As for the third model. While it is the model that I may appear to be most personally partial towards, I believe it also falls short when describing the complexity of the brain and of consciousness. I do not believe that tulpas and hosts are wholly separate all the time, all the way–as evidenced by things such as bleedover and thought confusion–though I do believe they are separate enough to possess their own largely distinct awarenesses and thus be considered separate entities in their own right like the host, especially when one takes functionality into the picture.

This leaves us at the second model, which I find simultaneously the most fascinating and potentially intricate, and yet also the most commonly oversimplified and oversimplifying. The reason for this is that it assumes a centrism about the host–that the unconscious workings of the brain “belong” to the host, that the unconscious is simply an extension of the conscious host, and thus all parts of the unconscious are simply parts of the host. And that thus, the host always takes priority not out of any pragmatic functionality, but simply because they are somehow intrinsically most important–and the unconscious and various parts of it are simply things to wrestle into submission.

That is an immense oversimplification, and a rather arbitrary one. Given the many processes in ourselves we are not aware of, that drive us regardless–the most predominant example being emotion–and a spiel about the illusory nature of free will that I won’t go into here, it is perhaps more accurate to say that hosts themselves are conscious extensions of the unconscious. If the unconscious “belongs” to them, it belongs not in the sense of a limb, but a stretch of half-wild property that must be adapted to and learnt from rather than passively commanded, and it “belongs” to all “extensions” of the unconscious, not merely only to the host. And each of these extensions, by virtue of possessing the same underlying mechanisms for processing and their own pool of memory, is as conscious as the host.

I have seen, in the DID community, many who take the “systemmates as parts” model and transform it into what’s very much like the “systemmates are people” view with different language. Their reasoning is roughly as follows: even if we are all “parts” of one person, we each are conscious because each “part” of a person is still conscious. We each have our own experiences separate from the other “parts”. Because we are each conscious with our own experiences, we are thus each equally real and conscious individuals. Consciousness is, to them, a fractal, not a solid monolith. In any case, it is more complicated than “they are just parts of the host”.

Perhaps the best way to explain this all would be through a metaphor. The first model asserts that the host is a lone tree, and tulpas are facades of trees arranged about facing the tree. The third asserts that the host is a tree in an orchard, where tulpas are other trees. The second asserts that the host is a tree, and that tulpas are that tree’s branches. I see it as such: there is a tree, but it is not the host. The host–their consciousness, their awareness–is a branch of that tree, and the tulpas are also branches–no single branch is the trunk of that tree. No one branch “belongs” to another one branch, unless all branches “belong” to each other. Each branch, though connected through the tree, is still its own branch, and though perhaps some branches may be thinner and lost with less consequence than others, each branch is still equally real, and capable of growing into something as large and strong as the “first” branch.

Now, again, that was all personal speculation on my behalf that I couldn’t resist adding. In the end, we still do not know how to even begin solving the hard question of consciousness, and we probably still know less about the brain than we do about Mars. The presence and dimensions of consciousness, whether in tulpas, animals, or AI, is still an unanswerable question.

However, there is one point at which everything can converge–and that is what we can observe about tulpas themselves. We can observe that they function like physical individuals, possessing the same lines of emotional and cognitive logic. Though we cannot see what is happening within the black box, we can observe that the black box does indeed possess sapience. Thus, we can apply the same logic we use when confronted with the possibility of other individuals being p-zombies–that regardless of what they “actually” are, we can apply our values in response to their behaviors in a pragmatic fashion.

We can observe that treating a tulpa well, allowing them independence and freedom of expression, and regarding them with respect leads to more rapid growth and differentiation than stifling. Positive treatment of a tulpa, as with physical individuals, leads to positive relationships. Conflicts can be resolved with communication rather than coercion, with both parties emerging the wiser for it. On the other hand, treating someone like a person does not mean allowing them to walk over you–if a tulpa is pushy, it is a host’s right to stand up for themself. If a tulpa is hostile, it is a host’s right to enforce the law of the land. If there are physical-world responsibilities to attend to, authority for that matter rests with the individual(s) taking direct management of that matter, whether the tulpa or the host. Beyond that, the specifics of how a host treats a tulpa will boil down to personal values, as it so often does with individuals in physical communities, and as much a source of disagreement it may be.

I suppose that if you had to take away one thing from this wall of text of a section, it’s that we don’t know and probably will never know in our lifespans what tulpas “actually” are. But regardless of what they “actually” are, when it comes to how they tick, they aren’t so different from hosts after all. And that, in essence, is the very point of tulpamancy.

Tulpamancy and Plurality

And now for something entirely different! With an upsurge in questions involving plurality, I thought I might include a section on how tulpamancy relates to it.

Simply put: plurality is an umbrella term for all experiences involving or relating to sharing a head with others. As the term implies, there are many more ways to experience head-sharing than tulpamancy and DID/OSDD. To list a few: soulbonding, natural/endogenic multiplicity, various religious experiences (hearing God, spirit possession around the world), arguably daemonism and voice-hearing as well.

The definition the broader plurality community uses for “tulpa” tends to differ somewhat from the definition I provided at the beginning of this very long post. In this community, it means “someone you share a head with who isn’t an alter”. In the broader spectrum of plurality, it tends to mean “someone you share a head with who you consciously (though not always deliberately) created in some way” or “a non-original headperson who considers themself a tulpa and of the tulpamancy community.” Indeed, there’s a few head-collectives who’ve stumbled on tulpamancy, and because it was the only thing they’d ever seen that remotely resembled what they had, they thought they were tulpamancer and tulpas. Later on, when informed of other experiences, they found out they fit better in multiple (clinical or endogenic), soulbonding, daemonism, or other spaces instead.

All in all, though, keep in mind that definitions and origins can be very blurry. The tulpamancy community for the longest time has been made up of people from all over the spectrum of origin, and all of them still were able to contribute insights everyone could benefit from. It’s just that people didn’t know that there was more out there than tulpamancy until recently, and thus used terms like “natural tulpa”, “accidental tulpa”, “fandom tulpa” instead of walk-in, soulbond, fictive, etc. Honestly, labels in plurality are less about origin and essence and more about where one identifies and what cultures they consider themselves part of. What separates the points on the spectrum is more culture and experience, than what their members “actually” are.

Or, if you’d like a metaphor: “culture” is an umbrella term for how people in societies across the world organize themselves and understand their world. Some of these cultures can be in direct conflict with each other, such as Pakistani and Indian culture, to the extent that mistaking one for the other can be a deadly offense–however, both are still very real cultures. And in many cases, boundaries between cultures are not absolute–there is a great deal of diffusion and cross-chatter, as we see with Westernization, and a great deal of people who straddle the line between cultures, as with Chinese-Americans and Chinese and American culture. What divides these cultures isn’t so much the biologies of their people, but their beliefs, practices, experiences, and worldviews. So too with the many points on the plurality spectrum, tulpamancy included.

Conclusion: Is Tulpamancy For Me?

I’ll end this by addressing the demographic that probably finds this all the most personally relevant–those considering making tulpas.

The simple way to put it: we can’t tell you whether or not you should make a tulpa. It’s a highly personal decision, and none of us are qualified to make it for you. If you are considering making a tulpa, I would recommend detailed introspection on why and how. Some questions to begin with:

Be brutally honest with yourself. Append “and why, and how?” to the end of each question you ask yourself.

You should also consider potential consequences. To list a few:

All in all, tulpamancy can be an extremely rewarding experience. Just be sure to give it some thought first–pun not intended. ;)

Thank you for reading. And to all tulpamancers young and old, old and new: happy forcing!

Q. Okay, that’s cool, but can you give me proof that tulpamancy is A Thing That Exists and not just a truly epic trolling method? A. To cut a long story short:

Excerpt from “The Illusion of Independent Agency : Do Adult Fiction Writers Experience Their Characters as Having Minds of Their Own” in Imagination, Cognition and Personality:

At first glance, the activities of an adult fiction writer might seem far removed from those of a child playing with an imaginary companion… However, when we surveyed accounts of the writing process, we were struck by the number of authors who described having personal relationships with their characters and imagined conversations with them.

For example, Francine de Plessis Gray described her characters as sleeping in her bed with her and sometimes waking her up to ask about her plans for their future. Alice Walker reported having lived for a year with her characters Celie and Shug while writing the novel The Color Purple. Walker writes, “Just as summer was ending, one or more of my characters—Celie, Shug, Albert, Sofia, or Harpo—would come for a visit. We would sit wherever I was, and talk. They were very obliging, engaging, and jolly. They were, of course, at the end of their story but were telling it to me from the beginning.” […]

In these accounts, writers describe their characters as autonomous beings who exist and act outside of their authors’ control and have minds of their own. They arrive fully formed in the mind’s eye and are resistant to change. For example, when J. K. Rowling […] was asked in a National Public Radio interview […] “I never write and say, ‘OK, now I need this sort of character.’ My characters come to me in this sort of mysterious process that no one really understands, they just pop up.” […]

Sometimes characters are described as having definite opinions about the narrative in which they live. They aregue with the author about the direction the novel is taking and their actions in it. […] Sara Paretsky described making a deal with her recurring character V. I. Warshawski. The story line in her novel Hard Time required that Warshawski go to prison where she was beaten up, tortured, and almost killed. V. I. “refused” to go along with this until the author promised to give her true love in exchange. Similarly, Philip Pullman […] described having to negotiate with a particularly proud and high strung character, Mrs. Coulter, to make her spend some time in a cave at the beginning of The Amber Spyglass.

In some accounts, the fictional characters do not limit their opinions to the world of the novel. They also provide unsolicited advice about matters concerning the author’s real life…

For those who aren’t familiar with Star Trek, Leonard Nimoy was the actor who portrayed Spock, a major character in the series. In his autobiographies I Am Not Spock and I Am Spock, he described having conversations with Spock in his mind, in which Spock took on a life of his own. He isn’t entirely clear on how independent Spock was or wasn’t (understandable, given that even an actor of his profile may have needed to fudge the details on something like this), nor is he clear on how many of the dialogues were actually had or were created for the book, but it’s something interesting to read, and another little reminder that this isn’t as outlandish or “insane” as the average layperson might think.

Here’s a few pages from the first chapter of I Am Spock: Page 1, Page 2, Page 3 and a bonus: the foreward to the book.