Promiseland

[Fetian]

So this happened!

  Hidden text. You are not a member of the groups to which this text is available

And I dunno man, from how much he struggles with the questions I just don't buy that he already knows everything! Also 'is there a way to take a more advanced class' is a question for a counsellor or a professor, not a random other student who's just asking why you're apparently wasting your time here.

That whole thing did get closer into bullying than I was comfortable with (including a bit I didn't include in screenshots where one of the students kind of doxxed disruptive student), but also man don't be having your breakdowns in the class chat, come on.

[Fetian]

Okay so! Obviously all of that ^ happened, but in actual class terms we went over te forms used for present tense and habituals, so the difference between "I eat" and "I'm eating". We went further into the book than I expected us to so I prepared for two out of the three exercises we ended up doing, and unfortunately that last one was pair work in a breakout room so someone else had to deal with me being very bad at everything. The teacher also jumped us right into grammar work and then halfway through class was like 'Oh I guess I should teach you the vocabulary for this chapter' so then we went back and went over the vocabulary.

After class I hung around to ask what I'd gotten wrong on the test -- I'd forgotten the "here" in "you're not allowed to take pictures here", so just something I missed and need to keep in mind going forward, not something I didn't know/understand and need to study more.

Immediately after class my audio drivers stopped working, so I tried restarting but my computer wouldn't do so, so I did a hard shutdown and things seem to be working fine now? Concerning nonetheless. I'd intended to do the Arcadia credits right after class but with that going on I ran out of spoons, so went off to eat dinner instead. That was at like 8 o'clock, so pretty late for that.

ash and I watched the latest Australia Taskmaster and then he went to bed and I went off to try to knock out the Arcadia credits before Stardew with Dave, but there was a minor issue in said credits so I put them off again, and Dave and I played Stardew for about three hours. I have revamped my basement, and we're still working on gathering walnuts for island upgrades. I don't think we accomplished much this week, just made general progress toward things.

Now I'm going to sit for a minute, then doublecheck the credits just to get a second set of eyes on them, since it's a sort of final thank-you to the cast and it would suck for someone to get left out, and then I will go to bed! Tomorrow I will have homework to do, and having taken a peek at it earlier it's going to be time consuming, and then I also plan on mopping the floor and finally finally getting the dresser moved. I also have been thinking about rearranging a couple other furniture pieces in here, because I think putting the cat tree by the window would be nice, but dresser and library table first and foremost.

[Fetian]

ough I locked myself into running a computer update that's taken almost two hours and counting

[Fetian]

I also have run out of stomach pills, and I swear I had more bottles somewhere but I can't find them

[Fetian]

ough I locked myself into running a computer update that's taken almost two hours and counting

Finally done

Criminy I'm tired

[Fetian]

Trapped under a sleeping cat

[Fetian]

See, this is "false fall"

Overcast, rain, getting below 60 at night. Impossible to fully wake up because I haven't seen the sun all day. And we're getting about a week of it, not just a day or two of 80F and some wind

And next week we're upper 90s again

[Fetian]

Woof, homework done, with twenty-five minutes to spare. Didn't do kanji this time because that always makes it take so much longer, hopefully that's not a problem.

[Fetian]

Somehow failed to write out part of one of my answers?? So rushed to get that redone and resubmitted, now six minutes before the deadline.

[Fetian]

So! Online translators and grammar-checkers. Decided to test out a few in an attempt to find one that will work for me, using these sentences:

花の写真を撮っています。
写真の花を撮っています。

Here's a breakdown:
花 = flowers
写真 = picture
とっています = taking, as in a picture
の = possessive particle
を = object particle

So, these sentences will definitely mean different things, and forgetting which word goes on which side of the possessive particle is the sort of thing that a real person would do and need help with, and is the sort of grammar error that any checker should be able to check. One of these is a grammatical sentence that means "I am taking pictures of flowers" and the other is ungrammatical and would mean something like "I am taking the picture's flowers (but not in the sense of possessing, in the sense of taking a picture)"

I've popped open every result for "check japanese grammar" on the first two pages of both duckduckgo and google -- let's see how they do

I'll start with google translate, which has predictable results if you've been paying attention:

Screenshot 2024-09-18 at 23-55-26 Google Translate.png

Screenshot 2024-09-18 at 23-56-48 Google Translate.png

If I don't use the kanji for 'take' (making it indistinguishable from the verb 'to take, as in into your possession'), it does try something different:

Screenshot 2024-09-18 at 23-58-47 Google Translate.png

Unless you include the grammatical sentence, which throws it off for some reason:

Screenshot 2024-09-18 at 23-59-58 Google Translate.png

So that's a fail as a grammar checker, which fair! Google translate is not a grammar checker! However it's also a failure at a translation, so. Zero marks.

 
Next, bunpo-check. All of these are going to be AI-based, so I'm not going to mention that every time -- I would love to have a tool that isn't a generative AI run on an LLM, but that doesn't exist right now as far as I can find. This one has a bonus of using what you put into it to train future models, so it also gets marked down for that, though it gets a very cynical point for actually telling you that's so. It also requires a sign-in and restricts you to 2000 characters a day.

It does not have a translation option, so you do have to just assume that what you've said is what you meant to say. This one does catch that the ungrammatical sentence is ungrammatical, but does not give a correct correction:

The correction from '写真の花をとっています。' to '写真を撮っています。' removes 'の花' (the flower) because it incorrectly suggests that the flower is taking the photo. The corrected sentence '写真を撮っています。' simply means "taking a photo," which is the intended meaning. The original sentence was grammatically awkward and did not clearly convey the action of taking a photo.

While the change does make it grammatical, it also removes information and doesn't clarify to the user that the issue is one of transposing which item goes on which side of the possessive marker (instead implying that the 花の is superfluous). I'm also not at all sure how the original would imply the flower is taking the photo, as it is still being marked by the object particle, so it seems to be pulling that entirely out of someone's ass.

 
Sapling AI. This one also doesn't have a translator, and claims that the sentences entered have 12 words in them, which is nuts. Even the romaji it shows at the bottom only gets up to 8 words.

Anyway, it does find an issue with the ungrammatical sentence, and does suggest replacing the 写 with 花の写 and removing the の花 -- which is a very confusing way to fix this but would work fine, except that on doing the first fix it then decides the sentence is grammatical and gets rid of the second suggestion. Leaving you with 花の写真の花を撮っています. Or, 'the flower's picture's flowers', which I'm taking (in the sense of taking a picture).

That one also doesn't work in firefox, so fuck it anyway.

 
Arvin. Also doesn't translate, and instead of a 'check grammar' button of any kind, it has 'optimize with AI'. It doesn't do anything to the grammatical sentence, as one would probably expect, but it turns the ungrammatical one into '私は写真の花を撮っています', or, uh... 'as for me, the picture's flower I'm taking', I guess. So, ungrammatical in exactly the same way but it's made it more explicit who the subject is?

 
Writecream is one I had some hopes for, despite its awful name -- rather than spit out a simple translation or suggested corrections, it instead appears to be prompting the AI to 'please explain the grammar of this sentence', and so outputs enormous paragraphs breaking everything down. It also does this several times for some reason, I can only guess that behind the scenes it is poking several different LLMs, which is probably unnecessary and wasteful. But here's some of what it has to say about the grammatical sentence:

The Japanese grammar in English for the text "花の写真を撮っています。" is: * 花 (hana): Flower * の (no): Possessive particle, indicating possession or belonging * 写真 (shashin): Photograph * を (o): Object marker * 撮る (toru): To take (a picture) * て (te): Te-form of the verb, indicating a continuous action * いる (iru): To be, indicating a state of being Breakdown: * 花の写真 (hana no shashin): "Flower's photograph" or "photograph of a flower" * を撮っています (o totte imasu): "I am taking [a photograph]." Therefore, the full translation is "I am taking a picture of a flower."

The Japanese sentence "花の写真を撮っています" can be broken down into the following grammatical components: 1. 花 (hana): This is the noun "flower" and is the direct object of the verb. 2. の (no): This is the genitive particle, indicating possession or belonging. It connects "flower" to the following noun. 3. 写真 (shashin): This is the noun "photograph" and is modified by "flower" through the genitive particle "no". So, it translates to "photograph of a flower". 4. を (o): This is the direct object marker. It indicates that "photograph of a flower" is the direct object of the verb. 5. 撮る (toru): This is the verb "to take (a picture)". It is in the te-form, which is used to indicate a continuous action. 6. て (te): This is the te-form of the verb "toru", used to connect the verb to the following auxiliary verb. 7. います (imasu): This is the present tense form of the auxiliary verb "iru". It indicates that the action is happening at the present moment. Overall Structure: The sentence structure follows the typical Subject-Object-Verb (SOV) order common in Japanese: * Subject: (Implied) "I" * Object: "花の写真 (photograph of a flower)" * Verb: "撮っています (am taking)" English Translation: "I am taking a picture of a flower."

Confusing, but thorough! Unfortunately, with the ungrammatical sentence, it:

The Japanese grammar in English for "写真を撮っています" is: * 写真 (shashin): Photograph (noun) * を (o): Object marker * 撮る (toru): To take (a picture) (verb) * て (te): Connects a verb to the following "いる" or "います" * いる (iru): To be (present progressive) Breakdown: The sentence structure is similar to English, but with different word order: * 写真 (shashin): This is the object of the sentence. * を (o): This particle marks the object. * 撮る (toru): This is the main verb, meaning "to take (a picture)". * て (te): This particle is used to connect the verb "toru" to the following "iru". * いる (iru): This verb indicates the present progressive tense. Therefore, the sentence "写真を撮っています" translates to "I am taking a picture" (or "I am taking photos").

-silently corrects the sentence by removing の花

The Japanese sentence "写真を花を撮っています" can be broken down as follows: * 写真 (shashin): This is the noun "photo" or "picture". * を (o): This is the object marker particle. It indicates that "photo" is the direct object of the verb. * 花 (hana): This is the noun "flower". * を (o): This is another object marker particle, indicating that "flower" is also a direct object. * 撮る (toru): This is the verb "to take (a picture)" * て (te): This is the "te-form" of the verb "toru", which is used to connect verbs and create a continuous action. * いる (iru): This is the verb "to be" used to indicate that the action is ongoing. Therefore, the sentence translates to "I am taking a picture of the flower" in English. This sentence showcases a common grammatical construction in Japanese, where verbs are often used with the "te-form" to express ongoing actions. The sentence is grammatically correct, but it's not the most natural way to say "I am taking a picture of a flower" in Japanese. A more natural sentence would be: 花の 写真を 撮っています。 (hana no shashin o totte imasu.) This sentence uses a slightly different word order, placing the noun "flower" first and then using the possessive particle "の" (no) to connect it to the noun "photo." This word order is more common and creates a smoother flow in the Japanese language.

-silently corrects the sentence to "写真を花を撮っています", which I believe is also ungrammatical, calls it 'unnatural', and then corrects that to the correct sentence (so at least it gets there)

The Japanese sentence "写真の花を撮っています" can be broken down for grammar analysis as follows: 1. 写真 (しゃしん - shashin): This noun means "photo" or "picture." It serves as a modifier for the noun that follows. 2. の (no): This particle indicates possession or a descriptive relationship. In this case, it shows that the flower is associated with the photo, translating to "of." 3. 花 (はな - hana): This noun means "flower." It is the subject of the action in this sentence. 4. を (wo): This particle marks the direct object of the verb. It indicates what is being acted upon—in this case, the flower. 5. 撮っています (とっています - totteimasu): This is the te-form of the verb 撮る (とる - toru), which means "to take" or "to photograph." The phrase "ています" (teimasu) indicates a continuous or ongoing action, translating to "am taking" in English. Full Translation: "I am taking a picture of the flower." This breakdown highlights how possession, action, and objects are structured in Japanese grammar compared to English.

-treats the ungrammatical sentence as grammatical -- I think it gets here by ignoring the 写真の and analysing the sentence as 花を撮っています -- which may be grammatical, I don't know. It is, however, not the sentence I asked it to analyse.

Notably none of this is what the writecream webpage says it's going to do.

How It Works:

1. Text Input: Users input their Japanese text into WriteCream’s interface, either by typing directly or pasting from another source.

2. Automatic Analysis: The tool automatically scans the text for grammatical errors, including incorrect verb conjugations, particle usage, and sentence structure issues.

3. Error Highlighting: It highlights the detected errors within the text, making them easily identifiable for the user.

4. Correction Suggestions: WriteCream provides accurate correction suggestions, offering alternative wordings and explanations for the errors.

5. Single-Click Fixes: With a single click, users can apply the suggested corrections, instantly refining their text to be grammatically accurate and professional.

Screenshot 2024-09-19 at 00-43-29 AI Japanese Grammar Checker 100_ Free No Login.png

No highlights, no suggestions, no fixes. Very strange.

 
Textgears. Finds no errors with the ungrammatical sentence, translates it as 'I am taking pictures of flowers'. I suspect it's piggybacking off google.

 
Languagetool. Finds no errors with the ungrammatical sentence, and doesn't have a translation option.

 
yeschat appears to be one of those LLM chatbots that people have set up to fulfil a specific purpose -- this one claims the sentence is grammatical and claims that it means 'I'm taking a picture of a flower that's already in a photograph' (hey maybe! I'm not an expert! Very suspicious though!)

That one locked me out of talking to it after three statements so I couldn't dig in any further.

 
Every other site I looked at either didn't work at all or wanted money, so let's round this off with what ChatGPT has to say about it.

You said:
Is the sentence "写真の花を撮っています" grammatical?

ChatGPT said:
Yes, the sentence "写真の花を撮っています" is grammatically correct in Japanese. However, its meaning might seem a bit unclear because of the combination of "写真" (photo) and "花" (flower). Here's a breakdown:

Code: Select all

写真 (しゃしん) = photo
の = possessive particle, which links two nouns (e.g., "of" in English)
花 (はな) = flower
を = object particle, indicating the direct object of a verb
撮っています (とっています) = the verb "to take (a photo)," in the progressive form, meaning "is taking"

So, the sentence literally translates as: "I am taking the flower of the photo," which sounds odd in English.

If you meant to say, "I am taking a photo of a flower," a more natural sentence would be:

「花の写真を撮っています」
This directly translates as "I am taking a photo of a flower," where "花の写真" means "a photo of a flower."

So, while the original sentence is grammatically correct, its meaning may not be what you intended.

You said:
Are you sure it's grammatical?

ChatGPT said:
You're right to question that! Upon closer examination, the sentence "写真の花を撮っています" is not grammatically sound, even though the structure looks correct at first glance.

The issue arises from the unnatural use of "写真の花" (photo's flower), which implies a possessive relationship that doesn't really make sense. This phrase would mean "the flower of the photo," which is awkward and unclear because a photo doesn't possess a flower.

To express the intended meaning of "I am taking a photo of a flower," the correct sentence should be:

「花の写真を撮っています」
This means "I am taking a photo of a flower," where "花の写真" (photo of a flower) makes sense grammatically and logically.

So to correct myself: "写真の花を撮っています" is not fully grammatical because of the odd relationship between "写真" and "花."

Classic.

 
So. In conclusion, you cannot trust LLMs to correct your grammar, even if they're not outright hallucinating. You also can't trust them to translate for you, because they will hallucinate your statements into things that make the most sense to them. And I'm just shit out of luck in finding a way to crosscheck my work.

At least there's clearly still job security in translation. Sucks that we're destroying the economy, the planet, and several industries for this though.