Rupak Goswami

The integration of Artificial Intelligence (AI) into writing assessment in Indian Higher Education (IHE) promises efficiency but raises critical questions about validity and ethics. This qualitative case study explores faculty perceptions of AI-based writing assessment tools. Grounded in a sociocultural theory of writing and Messick’s framework of consequential validity, this research analyses semi-structured interviews with 15 faculty members across four domains: physical, cognitive, information, and governance. Findings reveal that faculty view the uncritical implementation of AI as a significant threat to valid assessment. Key concerns include reinforcing the digital divide, cultural and linguistic bias against Indian English, the de-skilling of educators, and promoting shallow learning. These ethical and pedagogical issues are seen as fundamental challenges to the fairness and appropriateness of inferences drawn from AI-generated scores. The study concludes that for AI to be considered valid in the IHE context, its implementation must be preceded by robust institutional governance, a commitment to digital equity, and a design that respects sociocultural complexities. This research argues that ethical considerations are inextricably linked to the core principles of assessment validity.

Agriculture in the Indian Sundarbans deltaic region primarily depends on a rice-based monocropping system during the rainy season, with the subsequent season often remaining fallow. To mitigate this issue, a series of experiments using zero tillage and straw mulching (ZTSM) potato cultivation were conducted over eight consecutive years (2017–2024) across various islands in the Sundarbans Delta, West Bengal, aimed to intensify the cropping system and ensure the betterment of the land use pattern using climate-smart agricultural practices. In the initial two years, the experiments concentrated on assessing different potato cultivars and nutrient dosages under zero tillage and paddy straw mulching conditions. During the subsequent years, the focus shifted to field demonstrations under diverse climatic conditions. The research included the application of different macronutrients and growth regulators, in combination with different depths of straw mulching. In the final years of the study, the intervention was dedicated solely to the horizontal expansion of cultivated land. These initiatives aimed to enhance agricultural productivity and sustainable land use in the polders, promoting climate-resilient farming practices. From the sets of experiments, we standardized the sustainable nutrient management strategies and selection of appropriate potato cultivars vis-à-vis depth of straw mulching and, finally, the overall best agronomic practices for the region. The adoption of the ZTSM potato cultivation system demonstrated considerable success, as evidenced by the remarkable increase in the number of farmers employing this sustainable agricultural practice. The number of farmers practicing zero tillage potato cultivation surged from 23 in the initial year to over 1100, covering an area of more than 15 ha, highlighting the effectiveness of the technology. The analysis of the estimated adoption also showed that more than 90% adoption is likely to be achieved within a decade. This potential expansion underscores the benefits of the ZTSM potato cultivation system in improving soil health, conserving water, and reducing labour and costs. As more farmers recognize the advantages of zero tillage potato mulching, this approach is poised to play a pivotal role in sustainable agriculture, enhancing productivity while promoting environmental stewardship.

Upscaling sustainable intensification (SI) technologies is crucial to enhancing the resilience of fragile farming systems and vulnerable livelihoods of smallholder farmers. It is also critical to shape the future land-use and land-cover changes in a region. Zero-tillage potato cultivation (ZTPC), introduced as an SI intervention in parts of the Indian Sundarbans, has demonstrated promises of rapid upscaling, and thus, changes in the seasonal land-use pattern in the region. This study aims to understand the socioecological complexity of farming systems to comprehend how the nascent stage of ZTPC thrives at the farm level and what preconditions are necessary to upscale them. The objectives are to analyse the farm resource recycling pattern in ZTPC, and map and simulate its system’s complexity to strategize ZTPC upscaling in the region. The analysis of farm resource recycling data reveals that ZTPC stability hinges on managing trade-offs in resource allocations, specifically involving straw, organic manure, sweet water, and family labour. The decision to manage such trade-offs depends on farm type characterizations by their landholdings, distance from the homestead, pond, and cattle ownership, competing crops, and family composition. Using a semiquantitative systems model developed through fuzzy cognitive mapping, the study underscores the significance of effective training, input support, enterprise diversification by introducing livestock, timely tuber supply, access to critical irrigation, and capacity building of local institutions as the essential preconditions to sustain and upscale ZTPC. This research contributes a systems perspective to predict agricultural land use within technology transfer initiatives, providing insights into how farm- and extra-farm factors influence resource allocations for ZTPC. Public extension offices must understand the trade-offs associated with straw, organic matter, and harvested water and design differentiated supports for different farm types. The most compelling interventions to upscale ZTPC includes farm diversification by introducing livestock through institutional convergence, pragmatic agroforestry initiatives to enhance on-farm biomass and fuel production, building awareness and integrating alternative energy use to save straw and cow dung, building social capital to ensure access to sweet irrigation water, and developing and/or strengthening farmer collectives to ensure the supply of quality tuber and marketing of farm produce.

Efficient resource utilization in small-scale farms is crucial to achieving farm sustainability through endogenous mechanisms. However, the precise mechanisms to integrate farm resources to achieve farm sustainability are not very clear yet. By capturing the interaction among farm resources as a network phenomenon, we aimed to identify the discrete resource interactions (RIs) associated with higher farm sustainability in different farm types of Indian Sundarbans. First, we assessed the sustainability of 140 integrated farms using a synthesized assessment framework. Then, we considered four network motifs, namely linkage (a one-way link between two resources), reciprocal linkage (a two-way link between two resources), triad (three resources having closed interconnectedness), and the presence of a farm resource at the core of a network. Using RI network data of 140 farms and employing a graph theoretic approach we identified discrete network motifs (i.e., resource interaction) associated with highly sustainable farms in different farm types. We found a predominance of rice, vegetables and pond-based integration and identified 32 linkages, 11 reciprocal linkages, 21 triads, and three resources at the network core that occurred and co-occurred on highly sustainable farms, and thus critical to achieving farm sustainability. Further, multivariate analyses established that the properties of RI networks could explain farm sustainability significantly. We anticipate that sustainability in small-scale farms can be achieved by strategically designing new RIs on the farm. However, there may be limitations to such achievement depending on the nature of RI and the type of farm.

Identifying productive, profitable, and less risky cropping systems is pivotal for ensuring sustainable farm–based livelihoods in the context of climatic uncertainties and market volatility, particularly in many developing nations. Conventional field crop research often identifies the best or optimal solutions based on treatment replicates at a specific point in time without considering the influence of market volatility and climatic uncertainties. To address this gap, we conducted an assessment of productivity profitability and climate– and market–uncertainty-driven risk for eleven different rice-based cropping systems (eight existing and three potential systems) in the coastal region of Gosaba Block, West Bengal, India. Farmers’ observations of the best, typical, and worst seasonal yields and price data for the selected cropping systems over the last five to seven years were collected from fifty farm households. Irrespective of the scenarios, the rice–lathyrus systems, followed by rice–onion and rice–lentil systems, recorded the lowest rice equivalent yields and system yields. However, the highest rice equivalent yields and system yields were recorded for rice–chilli systems, followed by rice–tomato and rice–potato–green-gram systems. Per hectare, total paid–out cost (TPC) of rice–tomato systems was higher, followed by rice–chilli, rice–potato–green–gram, and rice–potato systems. However, irrespective of seasonal conditions (best, normal, and worst), rice–chilli systems gave a higher net return followed by rice–tomato and rice–potato–green–gram systems. The rice–fallow system recorded the lowest value for both parameters. Under the worst seasonal conditions, the rice–onion system gave a negative net return. Under all the scenarios, the rice–chilli system gave the highest benefit over cost, followed by the rice–tomato, rice–potato–green-gram, and rice–potato systems. The cumulative probability distribution (CDF) of per ha net income of the rice–tomato system showed first–degree stochastic dominance over other systems, implying that the system is economically the most profitable and less risky. Additionally, the CDF of net income per ha of the rice–chilli system showed second–degree stochastic dominance over the rest of the systems, indicating that the system is economically more profitable and less risky than other rice/non–rice cropping systems except for the rice–tomato system. Furthermore, the risk analysis results suggest that the likelihood of obtaining negative net income was nil for the selected cropping systems, except the rice–onion system had a slight chance (<1%) of providing a negative net return. Considering the productivity and economic viability (e.g., profitability and risk) of different rice–based systems, it is recommended to promote the adoption of the rice–vegetable systems, especially rice–tomato and rice–chilli from among the existing systems and rice–potato–green-gram systems from among the potential systems, for achieving sustainable intensification in these coastal saline tracts of the region.

Identifying productive, profitable and less risky cropping systems is important for sustaining farm-based livelihoods in the context of climatic uncertainties and market volatility in many developing nations. Reductionist field crop research identifies the best-bet solutions based on treatment replicates at a single point in time, which may not account for price instability under climatic uncertainties and volatile markets. Keeping this in mind, we estimated productivity, profitability and weather-related risk from eleven different rice-based cropping systems (eight existing and three potential systems) in the coastal region of West Bengal, India. Information on the crop management practices, yield and prices of the component crops of these eleven cropping systems, under ’best,’ ’normal’ and ‘worst’ situations (scenario), were collected through a ques-tionnaire survey on50 farms of Gosaba Block, West Bengal, India. Irrespective of the scenarios, the rice-lathyrus systems, followed by rice-onion and rice-lentil systems recorded the lowest rice-equivalent yields and system yields. However, the highest rice-equivalent yields and system yields were recorded for rice-chilli systems, followed by rice-tomato and rice-potato-green gram systems. Per hectare total paid-out cost (TPC) of rice-tomato systems was higher, followed by rice-chilli, rice-potato-green gram and rice-potato systems. However, irrespective of seasonal conditions (best, normal and worst), rice-chilli systems gave a higher net return followed by rice-tomato and rice-potato-green gram systems. The rice-fallow system recorded the lowest value for both parameters. Under the worst seasonal conditions, the rice-onion system gave a negative net return. Under all the scenarios, the highest B:C ratio was observed for rice-chilli, rice-tomato, rice-potato-green gram and rice-potato systems. The cumulative probability distribution of the rice-tomato system showed first-degree stochastic dominance over other systems and the rice-chilli system showed second-degree stochastic dominance over the rest of the systems. Only the rice-onion system had a small chance (< 1%) of a negative net return, while the rest of the cropping systems were highly likely to get a positive net return. Taking productivity, economics, and risk assessment of different rice-based systems into account, rice-vegetable systems, especially rice-tomato and rice-chilli among the existing systems and rice-potato-green gram system among the potential systems, can be recommended for sustainable intensification in these coastal saline tracts of the region. We also discuss additional socio-economic factors explored by in-depth in-terviews with the farmers, which might influence the adoption and upscaling of these cropping systems in the region.