The Tulpa/Host Writing Experiment

[TheCrawlingCreepypasta]

Overview

Hello and good evening, everyone!

This is a little experiment i'd like to execute which I call, "The Writing Experiment". The idea of this test is to have a tulpa and their respective host both write a page(Or less) of custom text. To clarify, both subjects will be writing out whatever first comes to their lovely mind. The purpose is to examine the differences and similarities between the host and their tulpa, through, of course, the writings.

General rules

There's really nothing big here; just firstly distinguish yourself from your tulpa, write at least 2 paragraphs. Grammer really isn't a priority. Also, please post below.

If anyone sees any problems with my execution, feel free to correct me, or bash me.

Ciao!

[Einulf]

To clarify, do you want pics of handwritten text, or do we just type it? I ask because our handwriting differs quite a bit between us.

[TheCrawlingCreepypasta]

To clarify, do you want pics of handwritten text, or do we just type it? I ask because our handwriting differs quite a bit between us.

 

I'm fine what either one, really.

Also, the handwriting is different among you and your tulpas? 'Cuz I wanna see that!

But still, whatever floats your battleship.

[Yakumo]

Question:

What methods are you planning to use for the comparison?

I am no expert in stylometry but I guess you'd need quite a lot of text material from both parties for a decent statistical analysis. Of course the more the better, however there likely is also a minimum word or letter requirement beyond which an algorithm will not produce any meaningful results. Just wanna say that 2 paragraphs might not be sufficient.

 

I'd suggest you also perform the test on yourself as a sort of control sample, write some paragraphs on different topics and see if your method is able show they came from the same author. I'd be interested in the success rate.

 

Btw, isn't there a stylometric test included in the Tulpa survey?

[sushi]

Btw, isn't there a stylometric test included in the Tulpa survey?

 

Yes there is, but I think it's pretty useless. The text sample is very small and I think that writing the text knowing that it will be analyzed influences the subjects to write as differently from each other as possible.

 

The last time Fenchurch took the survey, we copied every answer she gave on that survey and the one from a year earlier and compared them to each other and to a large sample of text from my PR.

 

Fenchurch's results were closer to each other than to mine -- but that might be a result of hers being short responses to survey questions, whereas mine was much longer and more expository.

[Guest Anonymous]

Melian: I like to write in pink or purple and put hearts all over everything. My host Davie doesn't do that.

 

Mistgod: Conclusive evidence? I think it is! Done. Research complete.

[Guest Anonymous]

http://pastebin.com/GHsGgNpi

[Einulf]

We have a small handwriting sample from each of us here. This was done in a bit of a rush (as evidenced by Dark's mistake), so the differences may not be quite as pronounced as usual, but I believe that a few distinct styles can be seen, particularly between Rand's writing and the rest of us.

 

Please not however, we are not a host and Tulpas, but are actually a system that falls somewhere between median and OSDD-1. This may play a large part in why we have differing styles. I am certain it influences Rand's writing, as he is a fictive (fictional introject, a.k.a soulbond) that remembers writing with fountain pens and quills, hence the flowing nature of his letters.

 

We must retire soon, as a new semester awaits us on the morrow and our sleep schedule is rather unstable as is, so we will most likely take part in the actual experiment during the absurdly long break between our classes tomorrow. I believe that this post could easily qualify for my submission, but I may write another in the actual document for simplicity's sake. -Aura

[TheCrawlingCreepypasta]

Question:

What methods are you planning to use for the comparison?

I am no expert in stylometry but I guess you'd need quite a lot of text material from both parties for a decent statistical analysis. Of course the more the better, however there likely is also a minimum word or letter requirement beyond which an algorithm will not produce any meaningful results. Just wanna say that 2 paragraphs might not be sufficient.

 

I'd suggest you also perform the test on yourself as a sort of control sample, write some paragraphs on different topics and see if your method is able show they came from the same author. I'd be interested in the success rate.

 

Btw, isn't there a stylometric test included in the Tulpa survey?

 

Thank you for your fine points; yes, large quantities of data would be wonderful! I'm mostly curious, however, in what the paragraphs are talking about instead of the style they're handwritten. (I apologize; i'm terrible at articulating my thoughts. ) I'm very interested in seeing the different mindsets of a host and their tulpa in their articulated thoughts/writings. I believe you thought I was comparing handwriting? Again, thanks for your fine points and I hope you join in this experiment. If you wanna ask any more questions, feel free to ask!

And to everyone who has participated in this so far, you have my personal gratitude!

[sushi]

I believe you thought I was comparing handwriting?

 

The stylometric test is more about word choice and writing style. It compares use of unusual words or unusual word combinations, sentence length, and things like that. In the past, techniques like that have been used to identify the writer of works of uncertain authorship.

 

For example, just this year, these techniques were used to identify Double Falsehood as being made up mostly of writing by William Shakespeare and John Fletcher. This means that it's likely an edited version of Cardenio -- Shakespeare's lost play, which has been missing for centuries.

 

But yeah, definitely not what you're looking for. Nothing to do with the actual subject of the writing.