Lok Sabha speeches – interpretation in 22 languages

How is the simultaneous interpretation in 22 languages is done in India’s Lok Sabha?

In a landmark move for linguistic inclusion, India’s Lok Sabha expanded its simultaneous interpretation services to cover all 22 official languages listed in the Eighth Schedule of the Constitution, a rollout finalized during the Winter 2025 session.

Testbook

This monumental task of real-time, two-way translation is achieved through a combination of upgraded hardware, specialized human interpreters, and advanced artificial intelligence:

1. Audio and Hardware Infrastructure

To handle the sheer volume of audio feeds, the Lok Sabha’s audio systems were significantly upgraded to process 22 distinct language channels simultaneously. Every seat in the chamber, as well as the press and visitor galleries, is equipped with a headset and a channel selector. This allows Members of Parliament (MPs) to simply dial into their preferred language and listen to the proceedings in real time, regardless of what language the current speaker is using.

Inter-Parliamentary Union+ 1

2. The Human Interpreter Pool

The backbone of the system is the Simultaneous Interpretation Service. The Lok Sabha expanded its team, putting candidates through rigorous auditions and practice sessions to build a large pool of highly skilled human interpreters capable of handling live parliamentary discourse.

Inter-Parliamentary Union

3. AI and Machine Learning Integration

To bridge gaps and enhance the speed and accuracy of the human interpreters, the Parliament integrated modern translation technology. Much of this AI-driven infrastructure and machine learning capability is being developed in collaboration with IIT Madras’s Human-Centered Artificial Intelligence Center and the Pravartak Technologies Foundation.

Testbook

The Result

Historically, MPs who wanted to speak in a regional language that wasn’t Hindi or English had to provide advance notice—often up to an hour—so the interpretation desk could prepare. Furthermore, translation was previously capped at 18 languages.

Slator+ 1

With the recent additions of Kashmiri, Konkani, and Santali completing the 22-language constitutional mandate, MPs can now spontaneously speak in their mother tongues. This setup ensures that representatives from rural or historically marginalized regions can articulate complex policy positions authentically, without being restricted by language barriers.

Testbook+ 1

So, basically it is done manually by human interpreters

It is actually a blend of highly skilled human professionals and advanced AI. While human interpreters are still the backbone of the system—ensuring nuance, cultural context, and accuracy—AI is playing a rapidly growing role.

Here is how the two work together in the Lok Sabha today:

The Human Element

The Lok Sabha Secretariat maintains a dedicated Simultaneous Interpretation Service. These are highly trained professionals who operate in soundproof booths. The human process involves a continuous, high-pressure loop:

cms.rajyasabha.nic.in

  • Listening: They listen to the speaker in the “source” language (e.g., Santali). Boostlingo
  • Processing: They instantly analyze and structure the message. knowledge-centre-translation-interpretation.ec.europa.eu
  • Reproducing: They translate and speak the message into the target language (Hindi or English) almost simultaneously. Boostlingo

Historically, if an MP wanted to speak in one of the regional languages, they had to provide advance notice so the specific human interpreter could be readied.

cms.rajyasabha.nic.in

The AI Revolution

To achieve the massive scale of translating 22 languages simultaneously, the government partnered with institutions like IIT Madras’s Centre for Human-Centric Artificial Intelligence. AI and machine learning are being utilized to:

Testbook

  • Assist Human Interpreters: AI tools help human interpreters in the booth by rapidly retrieving exact terminology, numbers, or technical jargon on the spot. knowledge-centre-translation-interpretation.ec.europa.eu
  • Process Regional Languages: Technologies like the Bhashini platform leverage advanced speech recognition and natural language processing (NLP). This is especially crucial for languages that have lacked digital representation, allowing for real-time machine translation between tribal languages and major Indian languages. PIB+ 1
  • Automate Documentation: AI is being aggressively pushed to translate parliamentary debates and paperwork into the 22 scheduled languages, significantly speeding up the legislative process and public accessibility. SuperKalam

In short, the human interpreters provide the necessary cultural and contextual accuracy, while AI provides the speed, digital assistance, and sheer processing power required to handle 22 languages simultaneously.

PIB

Election info changes quickly. Verify responses with official sources.

But how do the interpreters match the speed of the speakers, because they normally speak very fast and they often change topics

Simultaneous interpretation pushes the human brain to its absolute cognitive limits. To keep up with fast-talking politicians who frequently pivot between topics, interpreters do not attempt to translate word-for-word. Instead, their priority is to convey the overall meaning and keep up with the flow of speech.

montero-ls.com+ 1

To survive the relentless pace of a parliamentary debate, they rely on a set of highly specialized mental strategies and teamwork:

1. Décalage (Strategic Delay)

Instead of speaking the exact second the politician speaks, interpreters maintain a deliberate delay, known in the industry as décalage (lag time). They stay a few seconds or a sentence fragment behind the speaker. This delay allows them to hear enough of the sentence to understand its context and grammatical direction before they begin forming the sentence in the target language.

2. Anticipation

Because interpreters are experts in parliamentary procedures and the political topics of the day, they use context clues to predict how a speaker will finish a thought. By understanding the culture, the speaker’s political stance, and the language’s grammatical structure, they can often begin translating a sentence before the speaker has actually uttered the final verb.

3. Compression and Paraphrasing

When a speaker is talking at breakneck speed, word-for-word translation is impossible. Interpreters are trained to instantly identify the core message and strip away rhetorical fluff. They will alternately paraphrase, summarize, and edit on the fly to convey the core meaning concisely. They might also resort to acronyms to shorten long names or choose generic terms over highly specific ones to save precious seconds.

Capital Linguists+ 1

4. The “Salami” Technique (Chunking)

Politicians often speak in long, winding, run-on sentences. If an interpreter tries to maintain that exact same sentence structure in another language, they will quickly get tangled in grammar and fall behind. Instead, they use “chunking”—slicing one massive sentence into several shorter, punchy, grammatically simple sentences in the target language.

5. The Co-Pilot System (Booth Teamwork)

Simultaneous interpreting requires such intense concentration that an interpreter’s brain will physically fatigue if they go too long. Therefore, they never work alone.

  • Rotations: Interpreters work in pairs in a soundproof booth, rotating who is on the microphone every 30 minutes on average. Capital Linguists
  • Active Support: While one interpreter is speaking, the other is not just resting. The “off-mic” partner acts as a co-pilot. They actively listen and give support by jotting down difficult numbers, acronyms, or names on a notepad, looking up terminology, or feeding vocabulary to their partner. They maintain a speechless side conversation using meaningful looks, gestures, and notes. Capital Linguists

If a speaker is truly moving too fast or becoming incomprehensible, the interpreter’s brain must seamlessly adapt, managing the cognitive load by adjusting their strategy on the fly based on the speaker’s speed, accent, and topic complexity. Ultimately, if they miss a minor stylistic flourish, they let it go—the absolute priority is ensuring the core legislative argument reaches the rest of the chamber intact.

montero-ls.com