[Intro: Kali Uchis]
A tremendous killer with a flow to slay
Everyone's watching, but she couldn't care less
KAROL and Kali Uchis
The perfect combo to forget about that pain
[Chorus: Kali Uchis]
Uh (Oh, oh, oh, oh)
The girl is turned on, sticks to me like a tattoo (Like a tattoo)
I guarantee you that nobody is tougher than you (Nobody like you)
May God bless that ass that rubs, uh-uh
Like a tattoo, uh-uh
[Verse 1: Kali Uchis]
Once I turn it on, I never stop, stop (Never stop)
Be careful, I don't talk, I shoot (I shoot)
If you still don't know, let me make it clear, clear (Clear)
Playing with me always comes at a cost
Soft Reggaetón, bitten lips
Diamonds trailing down my belly button
More than one is already lost
A doll from a Tarantino movie (Give it to me, papi)
Go easy, you've been warned (Oh-oh)
It's very likely you'll get addicted to me (Oh-oh)
And if you want what's forbidden (Oh-oh)
I'll give it to you hard, darling, I'll punish you (Oh-oh)
[Chorus: Kali Uchis]
Uh (Oh, oh, oh, oh)
The girl is turned on, sticks to me like a tattoo (Like a tattoo)
I guarantee you that nobody is tougher than you (Nobody like you)
May God bless that ass that rubs, uh-uh
Like a tattoo, uh-uh
[Verse 2: Kali Uchis]
Look, I'm soft like honey and coconut
Always rich and sweet like corn arepas
And just with my look, she got all wet up
Your girlfriend goes crazy when I arrive (I arrive)
Maria, Jenny, Catalina, and Sonia
I love my Brazilians and my Colombians (Prr)
Dominicans, Puerto Ricans, I love my Mexicans
And tonight, I'm a lesbian, you make me feel like it
[Chorus: KAROL G, Kali Uchis]
Uh (Oh, oh, oh, oh; ay, papi)
The girl is turned on, sticks to me like a tattoo (Like a tattoo)
I guarantee you that nobody is tougher than you (Nobody like you; come get it)
May God bless that ass that rubs, uh-uh
Like a tattoo, uh-uh (Real Hard)
[Verse 3: KAROL G]
The baby is aggressive with that cute face (Hey)
She's well-established in her whole neighborhood
Short skirt and gistro leaning out of the sunroof
Her ass leaves everyone on mute
Strawberry gloss to bring it down (Bring it down)
Quietly so that no one knows (Ah)
Show me what you have there for me to try it
I'm already feeling hot, come and join me
Strawberry gloss to bring it down (Bring it down)
Quietly so that no one knows (Ah)
She undressed, and I couldn't stop looking at her
That tattoo on her back leaves me breathless
[Chorus: Kali Uchis, KAROL G, Kali Uchis & KAROL G]
Uh (Oh, oh, oh, oh)
The girl is turned on, sticks to me like a tattoo (Like a tattoo) I guarantee you that nobody is cooler than you (Nobody like you)
May God bless that ass that rubs, uh-uh Like a tattoo, uh-uh
[Outro: KAROL G, Kali Uchis]
Hey, mami, how good does that tattoo look on you
How far does it go? Show me, let me see
Jajaja Re-Re-Reggaetón
This is an edition of Time-Travel Thursdays, a journey through The Atlantic’s archives to contextualize the present, surface delightful treasures, and examine the American idea. Sign up here.
You can tell a lot about a cultural moment by the words it invents. New phenomena, products, social movements, and moods require new language, and an idea without a name is unlikely to stick. The job of a dictionary is to be responsive—but not too reactive—to these trends, to catalog the new ways people are talking, which of course is the new ways they’re thinking. (Among others this year: generative AI, girlboss, meme stock, doomscroll.) Language conjures moments, but it also creates them.
For about a decade starting in January 1987, this magazine’s back page belonged intermittently to Word Watch, a column by Anne H. Soukhanov. Soukhanov was then an editor of The American Heritage Dictionary, and Word Watch was a catalog of terms the dictionary’s editors were tracking for possible inclusion in upcoming editions, based on mentions in the press and pop culture—a sort of first pass at the linguistic infrastructure of tomorrow, an educated guess at how we might describe the unknowable future.
Now that we’re in the future those editors were guessing about, many of the column’s selections feel inevitable: infomercial, Astroturf, zine, NIMBY, ’roid rage, restorative justice. In January 1987, three and a half decades before we had girl dinner, the inaugural Word Watch had graze: “to eat various appetizers … as a full meal.” In October 1989, Soukhanov described in detail a new game called paintball, “dedicated players” of which were apparently called splatmasters, and in February 1991, she noted the rise of “precious language and luscious photographs used to depict recipes or meals,” which she called gastroporn (close enough). Two months later, a word to watch was canola, as in the seed that makes the oil that is almost definitely sitting in your kitchen right now, but of which, back then, “U.S. farmers [had] yet to commit themselves to extensive planting.”
Other times, Word Watch feels like a museum of bad ideas and forgotten trends, which of course is even more entertaining. In January 1988, Word Watch defined blendo as “a style of interior decoration that mixes hightech, Eurostyle, and antique furnishings into an integrated, individualistic whole.” In June 1989, there was halter-top briefs, which I regret to inform you is “a woman’s sleeveless upper garment constructed from men’s knitted, close-fitting briefs,” and which at least one fashion writer predicted would be soon be “‘seen on streets, in stores, and in shopping malls everywhere.’”
In April 1991, the column noted the possible rise of the washing emporium, “a coin-operated laundry incorporating such features as a bar, a restaurant, entertainment, a fax machine, mailboxes, a photocopier, a snack bar, a dining room, and a study area.” It cited as evidence Rutland, Vermont’s Washbucklers, whose owner was quoted in The Boston Globe and then in Word Watch saying that his business “may prove to be ‘the new social center of the ’90s.’” A personal favorite of mine is Skycar, a car-size aircraft that would purportedly fly at altitudes up to 30,000 feet and take off and land vertically, into a parking spot. It would sell for just under $1 million in 1992 dollars, but, Soukhanov noted, “the price is expected to drop as production volume increases.”
Committing a new word to the dictionary is a pretty strange act, when you think about it. So is making a magazine. Both are an attempt at describing the world at present, with only the evidence currently available, in the face of certain obsolescence. Every dictionary edition and every magazine issue is out of date shortly after it’s published; this is by design. These things are iterative, meant to be replaced by something better and newer, which of course will then be replaced, too—always never catching up. I like Word Watch for the same reason I like The Atlantic’s archives as a whole: It lays bare the messiness of trying to describe this big, weird, changing world. It’s open-minded about what the future might look like. It makes mistakes. It does its best.
Washbucklers, by the way, is still open. It has solar panels now.
Every day, millions of people start the day by posting a greeting on social media. None of them expect to be arrested for their friendly morning ritual.
But that's exactly what happened to a Palestinian construction worker in 2017, when the caption "يصبحهم" ("good morning") on his Facebook selfie was auto-translated as "attack them."
A human Arabic speaker would have immediately recognized "يصبحهم" as an informal way to say "good morning". Not so AI. Machines are notoriously bad at dealing with variation, a key characteristic of all human languages.
With recent advances in automated translation, the belief is taking hold that humans, particularly English speakers, no longer need to learn other languages. Why bother with the effort when Google Translate and a host of other apps can do it for us?
In fact, some Anglophone universities are making precisely this argument to dismantle their language programs.
Unfortunately, language technologies are nowhere near being able to replace human language skills and will not be able to do so in the foreseeable future because machine language learning and human language learning differ in fundamental ways.
How machines learn languages
For machine translation, algorithms are trained on large amounts of texts to find the probabilities of different patterns of words. These texts can be both monolingual and bilingual.
Bilingual training data comes in the form of human-translated parallel texts. These are almost always based on the standard version of the training language, excluding dialects and slang phrases, as in the example above.
Diversity is a characteristic of all human languages, but diversity is a problem for machines. For instance, "deadly" means "causing death" in most varieties of English, and that is what appears in the training data.
The Australian meaning of "excellent" (from Aboriginal English) puts a spanner in the works. If you input "Deadly Awards" into any translation app, what you'll get in your target language is the equivalent of "death-causing awards."
How machines store languages
The internal linguistic diversity of English, as of any other language, is accompanied by great diversity across languages. Each language does things differently.
Tense, number or gender, for example, need to be grammatically encoded in some languages but not in others. Translating the simple English statement "I am a student" into German requires the inclusion of a grammatical gender marking and so will either end up as "I am a male student" or "I am a female student."
Furthermore, some languages are spoken by many people, have powerful nation states behind them, and are well resourced. Others are not.
"Well resourced" in the context of machine learning means that large digital corpora of training data are available.
The lists of language options offered by automated translation tools—like the list of 133 languages in which Google Translate is currently available—erase all these differences and suggest that each option is the same.
AI speaks English
Nothing could be further from the truth. English is in a class of its own, with over 90% of the training data behind large language models being in English.
The remainder comes from a few dozen languages, in which data of varying sizes are available. The majority of the world's 6,000+ languages are simply missing in action. Apps for some of these are now being created from models "pre-trained" on English, which further serves to cement the dominance of English.
One consequence of inequalities in the training data is that translations into English usually sound quite good because the app can draw both on bilingual and monolingual training data. This doesn't mean they are accurate: one recent study found about half of all questions in Vietnamese were incorrectly auto-translated as statements.
Machine-translated text into languages other than English is even more problematic and routinely riddled with mistakes. For instance, COVID-19 testing information auto-translated into German included invented words, grammatical errors, and inconsistencies.
What machine translation can and can't do
Machine translation is not as good as most people think, but it is useful to get the gist of web sites or be able to ask for directions in a tourist destination with the help of an app.
However, that is not where it ends. Translation apps are increasingly used in high-stakes contexts, such as hospitals, where staff may attempt to bypass human interpreters for quick communication with patients who have limited proficiency in English.
This causes big problems when, for instance, a patient's discharge instructions state the equivalent of "Your United States was normal"—an error resulting from the abbreviation "US" being used for "ultrasound" in medical contexts.
Therefore, there is consensus that translation apps are suitable only in risk-free or low-risk situations. Unfortunately, sometimes even a caption on a selfie can turn into a high-risk situation.
We need to cultivate human multilingual talent
Only humans can identify what constitutes a low- or high-risk situation and whether the use of machine translation may be appropriate. To make informed decisions, humans need to understand both how languages work and how machine learning works.
It could be argued that all the errors described here can be ironed out with more training data. There are two problems with this line of reasoning. First, AI already has more training data than any human will ever be able to ingest, yet makes mistakes no human with much lower levels of investment in their language learning would make.
Second, and more perniciously, training machines to do our language learning for us is incredibly costly. There are the well-known environmental costs of AI, of course. But there is also the cost of dismantling language teaching programs.
If we let go of language programs because we can outsource simple multilingual tasks to machines, we will never train humans to achieve advanced language proficiency. Even from the perspective of pure strategic national interest, the skills to communicate across language barriers in more risky contexts of economics, diplomacy or health care are essential.
Languages are diverse, fuzzy, variable, relational and deeply social. Algorithms are the opposite. By buying into the hype that machines can do our language work for us we dehumanize what it means to use languages to communicate, to make meaning, to create relationships and to build communities.
Provided by The Conversation
This article is republished from The Conversation under a Creative Commons license. Read the original article.
Citation: 'Your United States was normal': Has translation tech really made language learning redundant? (2023, November 22) retrieved 25 November 2023 from https://ift.tt/AWzucsM
This document is subject to copyright. Apart from any fair dealing for the purpose of private study or research, no part may be reproduced without the written permission. The content is provided for information purposes only.
[Intro: Kali Uchis]
A tremendous killer with a flow to slay
Everyone's watching, but she couldn't care less
KAROL and Kali Uchis
The perfect combo to forget about that pain
[Chorus: Kali Uchis]
Uh (Oh, oh, oh, oh)
The girl is turned on, sticks to me like a tattoo (Like a tattoo)
I guarantee you that nobody is tougher than you (Nobody like you)
May God bless that ass that rubs, uh-uh
Like a tattoo, uh-uh
[Verse 1: Kali Uchis]
Once I turn it on, I never stop, stop (Never stop)
Be careful, I don't talk, I shoot (I shoot)
If you still don't know, let me make it clear, clear (Clear)
Playing with me always comes at a cost
Soft Reggaetón, bitten lips
Diamonds trailing down my belly button
More than one is already lost
A doll from a Tarantino movie (Give it to me, papi)
Go easy, you've been warned (Oh-oh)
It's very likely you'll get addicted to me (Oh-oh)
And if you want what's forbidden (Oh-oh)
I'll give it to you hard, darling, I'll punish you (Oh-oh)
[Chorus: Kali Uchis]
Uh (Oh, oh, oh, oh)
The girl is turned on, sticks to me like a tattoo (Like a tattoo)
I guarantee you that nobody is tougher than you (Nobody like you)
May God bless that ass that rubs, uh-uh
Like a tattoo, uh-uh
[Verse 2: Kali Uchis]
Look, I'm soft like honey and coconut
Always rich and sweet like corn arepas
And just with my look, she got all wet up
Your girlfriend goes crazy when I arrive (I arrive)
Maria, Jenny, Catalina, and Sonia
I love my Brazilians and my Colombians (Prr)
Dominicans, Puerto Ricans, I love my Mexicans
And tonight, I'm a lesbian, you make me feel like it
[Chorus: KAROL G, Kali Uchis]
Uh (Oh, oh, oh, oh; ay, papi)
The girl is turned on, sticks to me like a tattoo (Like a tattoo)
I guarantee you that nobody is tougher than you (Nobody like you; come get it)
May God bless that ass that rubs, uh-uh
Like a tattoo, uh-uh (Real Hard)
[Verse 3: KAROL G]
The baby is aggressive with that cute face (Hey)
She's well-established in her whole neighborhood
Short skirt and gistro leaning out of the sunroof
Her ass leaves everyone on mute
Strawberry gloss to bring it down (Bring it down)
Quietly so that no one knows (Ah)
Show me what you have there for me to try it
I'm already feeling hot, come and join me
Strawberry gloss to bring it down (Bring it down)
Quietly so that no one knows (Ah)
She undressed, and I couldn't stop looking at her
That tattoo on her back leaves me breathless
[Chorus: Kali Uchis, KAROL G, Kali Uchis & KAROL G]
Uh (Oh, oh, oh, oh)
The girl is turned on, sticks to me like a tattoo (Like a tattoo) I guarantee you that nobody is cooler than you (Nobody like you)
May God bless that ass that rubs, uh-uh Like a tattoo, uh-uh
[Outro: KAROL G, Kali Uchis]
Hey, mami, how good does that tattoo look on you
How far does it go? Show me, let me see
Jajaja Re-Re-Reggaetón
The Google Pixel Buds Pro and Pixel Buds A-Series are some of our favorite wireless earbuds. They're intuitive and comfortable with strong audio quality and Google Assistant support. But a lot of features aren't obvious, including real-time translation.
Google's powerful translation tools are available on recent Pixel phones, but a pair of Pixel Buds means you don't need your phone to understand other languages. We show you what this translation feature can do, how to use it, and things to be aware of.
What you need to use Google Translate with Google Pixel Buds
Google Translate is available on all Pixel Buds, but you need the following to use Google Translate:
A Google-Assistant-compatible Android device running Android 6.0 or later
The latest version of the Google app.
The latest version of the Google Translate app.
Your Pixel Buds can help you:
Translate conversations in real time.
Translate and transcribe audio.
How to activate Conversation Mode with your Pixel Buds
Google's conversation translation feature, conversation mode, can be activated with or without Google Assistant. It doesn't matter which method you use to activate it. We show you the steps for both methods.
How to turn on conversation mode with Google Assistant
Even if you activate conversation mode with your Pixel Buds, you still need your phone to handle the translation processing.
Tap and hold either earbud or say "Hey Google" to activate your Google Assistant.
Say, "Help me speak <language>."
The Google Translate app opens in conversation mode on your device.
You're ready to talk. Your Pixel Buds give you a brief tutorial on using Google Translate in conversation mode. We walk you through the steps in detail later in this guide.
How to activate conversation mode without Google Assistant
You need Google Translate on your phone regardless of your chosen method. You may find it easier to skip talking to your earbuds and activate conversation mode from the Google Translate app. You don't need to take your earbuds out. They'll still pick up your voice and output translations.
Open the Google Translate app.
Tap the language name in the lower-left corner.
Select the language you speak.
Tap the language name in the lower-right corner.
Select the language spoken by the other person.
Tap Conversation in the lower-left corner of your screen.
How to translate a conversation with Pixel Buds
In conversation mode, Google Translate displays your speech in your language and the other person's language on-screen and repeats the translation in your earbuds.
Tap and hold your earbud.
Start talking while holding your finger on your earbud.
Remove your finger when finished.
The other person can tap and hold the Microphone button underneath their language to respond in kind.
How to translate and transcribe audio with Pixel Buds
You can translate and transcribe audio with or without Google Assistant, like in conversation mode. We recommend using this feature when listening to a single person continuously talking, like in a speech or a news report.
How to transcribe audio with Google Assistant
Tap and hold either earbud or say "Hey Google" to activate your Google Assistant.
Say, "Help me understand <language>."
The Google Translate app opens in transcribe mode on your device.
Tap Transcribe.
How to transcribe audio without Google Assistant
Open the Google Translate app.
Tap the Microphone button at the bottom of your screen.
Tap Transcribe.
How to use transcription mode in Google Translate
Transcription mode transcribes audio on your phone.
Tap the language button in the lower-left corner of your screen.
Select your spoken language.
Tap the language button in the lower-right corner of your screen.
Select the speaker's language.
Your Pixel Buds translate the audio in real time if it is a supported language pair.
Tips for using Google Translate with your Pixel Buds
The steps above get you started translating languages with your Pixel Buds, but you should do a few things to make the most of this feature.
Download languages for offline translation
You don't want to be caught out without an internet connection and be unable to understand anything around you. When selecting a language in Google Translate, tap the Download icon to the right of the language to save it to your phone.
Avoid using transcribe mode with multiple speakers
Transcribe mode isn't always reliable. It's best when only one speaker can be clearly heard.
Adjust touch and hold settings on your Pixel Buds
By default, your Pixel Buds are set up to activate Google Assistant when you tap the left earbud and to toggle ANC (on the Pixel Buds Pro) when you tap the right. Open the Pixel Buds app and adjust the Touch Control settings if this doesn't suit you.
Translate conversations in real time with your Pixel Buds
While your Pixel Buds don't do the heavy lifting when translating speech, you get the benefit of clear audio and a microphone near your mouth. If you're unimpressed by the sound quality of your earbuds, consider checking to see if you're getting the best fit possible, as this can make a difference.
This is an edition of Time-Travel Thursdays, a journey through The Atlantic’s archives to contextualize the present, surface delightful treasures, and examine the American idea. Sign up here.
You can tell a lot about a cultural moment by the words it invents. New phenomena, products, social movements, and moods require new language, and an idea without a name is unlikely to stick. The job of a dictionary is to be responsive—but not too reactive—to these trends, to catalog the new ways people are talking, which of course is the new ways they’re thinking. (Among others this year: generative AI, girlboss, meme stock, doomscroll.) Language conjures moments, but it also creates them.
For about a decade starting in January 1987, this magazine’s back page belonged intermittently to Word Watch, a column by Anne H. Soukhanov. Soukhanov was then an editor of The American Heritage Dictionary, and Word Watch was a catalog of terms the dictionary’s editors were tracking for possible inclusion in upcoming editions, based on mentions in the press and pop culture—a sort of first pass at the linguistic infrastructure of tomorrow, an educated guess at how we might describe the unknowable future.
Now that we’re in the future those editors were guessing about, many of the column’s selections feel inevitable: infomercial, Astroturf, zine, NIMBY, ’roid rage, restorative justice. In January 1987, three and a half decades before we had girl dinner, the inaugural Word Watch had graze: “to eat various appetizers … as a full meal.” In October 1989, Soukhanov described in detail a new game called paintball, “dedicated players” of which were apparently called splatmasters, and in February 1991, she noted the rise of “precious language and luscious photographs used to depict recipes or meals,” which she called gastroporn (close enough). Two months later, a word to watch was canola, as in the seed that makes the oil that is almost definitely sitting in your kitchen right now, but of which, back then, “U.S. farmers [had] yet to commit themselves to extensive planting.”
Other times, Word Watch feels like a museum of bad ideas and forgotten trends, which of course is even more entertaining. In January 1988, Word Watch defined blendo as “a style of interior decoration that mixes hightech, Eurostyle, and antique furnishings into an integrated, individualistic whole.” In June 1989, there was halter-top briefs, which I regret to inform you is “a woman’s sleeveless upper garment constructed from men’s knitted, close-fitting briefs,” and which at least one fashion writer predicted would be soon be “‘seen on streets, in stores, and in shopping malls everywhere.’”
In April 1991, the column noted the possible rise of the washing emporium, “a coin-operated laundry incorporating such features as a bar, a restaurant, entertainment, a fax machine, mailboxes, a photocopier, a snack bar, a dining room, and a study area.” It cited as evidence Rutland, Vermont’s Washbucklers, whose owner was quoted in The Boston Globe and then in Word Watch saying that his business “may prove to be ‘the new social center of the ’90s.’” A personal favorite of mine is Skycar, a car-size aircraft that would purportedly fly at altitudes up to 30,000 feet and take off and land vertically, into a parking spot. It would sell for just under $1 million in 1992 dollars, but, Soukhanov noted, “the price is expected to drop as production volume increases.”
Committing a new word to the dictionary is a pretty strange act, when you think about it. So is making a magazine. Both are an attempt at describing the world at present, with only the evidence currently available, in the face of certain obsolescence. Every dictionary edition and every magazine issue is out of date shortly after it’s published; this is by design. These things are iterative, meant to be replaced by something better and newer, which of course will then be replaced, too—always never catching up. I like Word Watch for the same reason I like The Atlantic’s archives as a whole: It lays bare the messiness of trying to describe this big, weird, changing world. It’s open-minded about what the future might look like. It makes mistakes. It does its best.
Washbucklers, by the way, is still open. It has solar panels now.
Since the first translation of the Bhagavad Gita in 1789 by Charles Wilkins, there have been hundreds of translations into English. Some have translated the poetry of the Gita into prose, and some others have attempted metrical poetry. Whereas prose translations do not convey the delights of the original, good metrical translations have no choice but to compromise on fidelity. This is also why commentaries exist, but, they tend to be loaded with terminology and digressions, and however useful, deter the eager reader from a continuous reading of the text.
Reading the Gita is a happy afternoon project for a reader who knows Sanskrit; for a reader who does not know Sanskrit, old-style translations may not offer the same easy reading speed. The Gita is a teaching, it is not a pedantic text. Thought-provoking wordplay, connections and associations between words, and implied meanings are present throughout the seven hundred verses. Sometimes, the meaning is adequately conveyed only when you take in the suggestiveness and resonance (dhvani), characteristics of memorable poetry.
As languages, Sanskrit and English are different from each other – not just in terms of the conceptual worlds they rise from and refer to, but also in their linguistic structures, their ‘deep grammar.’ Sanskrit nouns are declined, and thus, have case endings; therefore, in a sentence, the parts of speech are made clear within the words. A simple example – in the sentence “A girl goes to school”, “girl” is the subject, and “school” is the object. In Sanskrit, the noun “girl” will be in the nominative case (prathamā vibhakti), and the noun “school” will be in the accusative case (dvitīyā vibhakti). Therefore, we may structure the sentence any which way, including “to school – goes – girl”. Now if all those components of the Sanskrit sentence (to school / girl / goes/) were written separately, in different corners of a single page, they would still make sense. In Sanskrit, we comprehend the words along with their roles in the sentence. In a Sanskrit stanza, then, what decides the sequence of words is rhythm, smoothness, or metrical considerations.
This is also why when we read Sanskrit poetry, we may begin with analysing the “prose order” (anvaya, or connections). This involves a reordering of the words in the stanza in a way that becomes easier to follow.
Such structural freedom inherent in Sanskrit composition opens up possibilities in translation via contemporary poetics. Like science, poetics has also become more developed, and today there are more resources available to writers and readers. Poetry was once an oral art, and belonged to the dimension of time. Even though manuscripts and inscriptions were used, the dimension of space arrived into everyone’s daily lives through the printing press and the printed page. Translating ancient texts is also a translation from a temporal, to a temporal and spatial medium. The line, for example, does not need to hang on for life to the left margin. Spacing and connections are available to be used if they help understanding. Here is an example:
In my translation, the has been deconstructed into the components and relationships between them. The vedantic idea of “neti” (‘not this’) is imitated. i.e., how discarding reveals the true you. You discard and you discard (what you are not) and that is how you find out who you are. It is by discarding (whatever is intransient) that “you are”. The analogy between old clothes and the old worn-out body is stacked vertically, so it is a quick and easy connection upon the page.
At this stage, Krishna is only speaking about the nature of ātman – it is that which has the body. Readers will notice that Krishna does not tell Arjuna directly – “you are ātman.” Krishna explains the concept of ātman, and without saying so, lets Arjuna realize his identity with ātman. What occurs here is an identity shift, a point that is communicated without being spelt out. Hence, my translation uses vertical as well as horizontal space, and presents the sub-text vertically – “this is who you really are.” The reader can take it in visually, even as she continues to read the line.
On the other hand, if something is really emphatic in a text, it can be spatially demarcated from the other words in the verse. Here is another example from my translation:
Arjuna uses the verb “see” – (I see, paśyāmi) – in every single stanza from stanza 11.14 to 11.19 11.19. Noticing this makes one much more aware that what is being described is a spectacle and an extraordinary sight. This is not just a description of Krishna’s viśvarūpam (universal form) – it is a description of viśvarūpa-darśaṇam (the vision of Krishna’s universal form). Hence, I extracted the verb “see” into a separate column, and relocated everything that was seen within its frame.
Excerpted with permission from Bhagavad Gita: God’s Song, translated from the Sanskrit by Mani Rao, HarperCollins India.