Personal Pronouns in Japanese

Personal Pronouns in Japanese

I sometimes spend several hours a day reading news reports and official bulletins in English for my work and now, half of it as a hobby, and when I start using machine translation software to get an overview, I sometimes wonder if there are no "personal pronouns" in Japanese.

When translating English into Japanese, the subject must be omitted or selected so that it becomes natural Japanese, or else the translation will remain uncomfortable. This is further emphasized when using translation software.

To put it bluntly, if you do not know what you are trying to say in Japanese, you will not know what you are saying in English translation either. In other words, writing ability is being questioned, and it is obvious to Japanese speakers that grammatically correct Japanese here is not "correct Japanese.

Furthermore, these days, it seems that more and more Japanese speakers are speaking strange Japanese with "personal pronouns" in their normal speech, perhaps influenced by the English language.

To begin with, Japanese people do not use "you" as a "personal pronoun" in the same way as "you" in English. This is because there are very few conversations in which "anata" can be used naturally, and it is common to refer to the other person by his or her name, job title, or position.
This may be because there is no grammatical pronoun in Japanese that can be distinguished from a noun and is used as a "personal pronoun.

With this thought in mind, I did a little research.
Then, I found that specialists have already discussed various issues.

Instead of trying to come up with a poor imagination, a little research will give you an answer, even if it is not a definite one. I guess that is what studying is all about.

What does the translation software think?
I asked the chat GPT.
The following quote is a reference example, as I think a lot can change depending on how you ask the question, but I think it is a thought-provoking answer.

● The treatment of personal pronouns is certainly one of the most difficult issues in translating between Japanese and English.
In English, it is common to always explicitly state the speaker or object regardless of context (e.g., "I," "you," "he," "she," "it," "we," "they," etc.).
In Japanese, however, it is very common to omit subjects and personal pronouns in some contexts, and one is expected to infer them from the context.

In the development of translation software (e.g., Google Translate and DeepL) and AI language models (e.g., OpenAI's GPT-3), the following strategies are used to address this issue

1 Understanding context: modern AI models have the ability to "understand context". That is, they take into account the sentences before and after and infer what the omitted subject is. For example, "Yesterday, I went to a park. I took a lot of pictures there." AI understands that the subject of the second sentence is the same as the first sentence.

2 Learning from large amounts of data: The AI model learns from large amounts of Japanese and English sentences and the relationships and patterns between them. This allows them to understand how to handle omitted subjects in specific contexts.

3 Use neural networks: Neural networks, on which AI is based, have the ability to mimic the way humans understand and use natural language. This allows it to generate more natural translations.
Even so, it is not perfect and occasionally makes mistakes. This is a difficult part of machine translation and a challenge for future research and development.