Tech Collocations — 工程技术英文搭配
本质 技术语境有自己的 collocation 集 —— "perform an experiment / conduct a measurement / observe a behavior / apply a load / induce a current / mitigate noise" —— 选词错就显得 non-native 即便语法对。这套词汇是 datasheet / paper / app-note 的**"母语"。中国工程师写技术文档最常踩的坑**: ① 用日常词("we made an experiment"应是"we conducted");② 用错动词强度("show evidence of" vs "demonstrate");③ 多层 of 嵌套("the strategy of management of thermal of the system" 应是"the system's thermal management strategy")。本页围绕"30 个高频 verb-noun + 学术高级动词区分 + noun chain 替代 of + 权衡句型 + 研究语境词汇 + 反模式"6 个维度。
学习目标
读完本页后,你应该能够:
- 掌握 30 个高频 verb-noun pair(perform / conduct / measure / observe / apply / induce / mitigate / characterize / verify / validate)
- 区分 "characterize / parameterize / specify" 的实际用法差别
- 使用 noun chains("thermal management strategy" / "low-side gate driver topology")替代多层 of
- 写"权衡"句型("trade-off between X and Y / at the cost of Z / in exchange for W")
- 区分 implication / hypothesis / assumption / conclusion 的研究语境
- 识别 5 个 tech writing 反模式
1. 30 个高频 Verb-Noun Pair
技术写作 verb-noun 搭配集中在 30 个——这 30 个覆盖 80% 的 paper / datasheet / app-note 高频场景。
1.1 实验类(8 个)
实验 / 测试场景用专门的高频动词——日常词("do / make")显得 non-academic。
| 搭配 | 用法 |
|---|---|
| conduct an experiment | We conducted an experiment to verify... |
| carry out a measurement | Measurements were carried out at 25°C... |
| perform a test | A series of tests were performed... |
| run a simulation | We ran SPICE simulations using... |
| gather data | Data was gathered over a period of... |
| observe a behavior | The expected behavior was observed when... |
| record a value | The value was recorded at 10 ms intervals |
| calibrate an instrument | The probe was calibrated against a reference... |
1.2 分析类(6 个)
分析 / 研究场景用 academic 动词——避免"think about / look at"等口语化表达。
| 搭配 | 用法 |
|---|---|
| analyse the data | The data was analysed using... |
| investigate the cause | We investigated the root cause of... |
| examine the impact | This section examines the impact of... |
| study the behavior | We studied the behavior under... |
| assess the performance | The performance was assessed against... |
| evaluate the design | We evaluated 3 design alternatives... |
1.3 控制 / 操作类(8 个)
电路 / 系统操作用专门的工程动词:
| 搭配 | 用法 |
|---|---|
| apply a voltage / current / load | A 100 V step was applied to... |
| induce a current | The changing flux induces a current... |
| trigger a pulse | The MCU triggers a 1 us pulse... |
| drive a load | The driver drives a 5 nF capacitive load |
| regulate the output | The LDO regulates the output to 3.3 V |
| suppress noise | The filter suppresses out-of-band noise |
| mitigate ringing | We added a snubber to mitigate ringing |
| isolate the fault | DESAT detection isolates the fault... |
1.4 验证类(4 个)
验证 / 确认场景用 verify / validate / confirm / establish 4 个核心动词——这 4 个之间也有 nuance 区别(verify 是检验设计、validate 是检验模型、confirm 是确认假设、establish 是建立基准)。
| 搭配 | 用法 |
|---|---|
| verify the design | The design was verified through... |
| validate the model | The model was validated against measurements |
| confirm the hypothesis | The data confirms our hypothesis that... |
| establish a baseline | A baseline was established under... |
1.5 表征类(4 个)
表征类 4 个动词容易混淆(详见 §2)——但搭配各自固定:characterize 配 device、parameterize 配 model、specify 配 requirement、define 配 boundary。
| 搭配 | 用法 |
|---|---|
| characterize the device | The device was characterized using... |
| parameterize the model | We parameterized the model with... |
| specify the requirement | The standard specifies that... |
| define the boundary | The SOA defines the operational boundary |
2. Characterize / Parameterize / Specify 区别
这 3 个词中文都译"指定 / 描述",但英文用法严格不同。误用是中国工程师常见错误。
| 词 | 含义 | 用法场景 |
|---|---|---|
| characterize | 测得 / 描述设备的实际特性 | "We characterized the gate driver under various loads." |
| parameterize | 给模型确定参数值 | "The SPICE model was parameterized using measurement data." |
| specify | 定义要求 / 规格 | "The standard specifies a maximum of 50 mV ripple." |
注意:
- characterize 强调测量真实情况(从 device 中获取数据)
- parameterize 是把 characterize 的结果填进 model
- specify 是规定要求(写在 datasheet / standard 里)
例: "We characterized the SiC MOSFET under high-temperature conditions, then parameterized the SPICE model accordingly. The model satisfies the specifications in IEC 61800-3."
3. Noun Chain 替代多层 of
英文技术语篇大量用 noun chain 替代"X of Y of Z of..."嵌套。中国工程师习惯用 of 串联,显得啰嗦。
3.1 改写规则
把 of 嵌套改成 noun chain 的核心规则——倒过来写,最后一个 noun 是 head,前面都是修饰。
[adjective] + [modifier noun] + [head noun]
把多层 of 倒过来:最后一个 noun 是 head,前面的 noun 都是修饰。
3.2 实战对比
下面 8 组展示典型的 of 嵌套与 noun chain 改写——每组都把 4-5 层 of 压缩为 3 层 noun chain,字数减半信息不变。
| 啰嗦版(多层 of) | 简洁版(noun chain) |
|---|---|
| the strategy of management of thermal of the system | the system's thermal management strategy |
| the topology of the driver of the gate of the low-side | the low-side gate driver topology |
| the analysis of the impact of switching on the efficiency | the switching efficiency impact analysis |
| the interface of communication between the MCU and the SBC | the MCU-SBC communication interface |
| the design of the package of the module of power | the power module package design |
| the algorithm of control of the loop of current | the current loop control algorithm |
| the behavior of the device under the conditions of high temperature | the high-temperature device behavior |
| the requirement of safety of functional of the system | the system's functional safety requirement |
3.3 noun chain 的限制
不要堆超过 4 层——超过 4 层就变成 jargon hell。3 层是 sweet spot:
- ❌ 5 层: high-temperature high-voltage SiC MOSFET reliability characterization
- ✅ 3 层: SiC MOSFET reliability characterization
- ✅ 拆: high-voltage SiC MOSFET reliability characterization at high temperature
4. 权衡(Trade-off)句型
技术写作最常出现的逻辑关系是权衡 —— X 提升带来 Y 下降。5 种权衡句型让你精准表达 trade-off。
4.1 5 种句型
下面 5 种权衡句型按"中性 → 强烈"递增——选择哪一种取决于你想强调的代价 / 收益的强度。
| 句型 | 例 |
|---|---|
| trade-off between X and Y | There is a fundamental trade-off between switching speed and EMI. |
| at the cost of Y | We achieved 10% efficiency gain at the cost of double the BOM. |
| in exchange for Y | We accept higher complexity in exchange for tighter regulation. |
| sacrifice X for Y | The design sacrifices flexibility for thermal robustness. |
| X comes at the expense of Y | Faster switching comes at the expense of EMI compliance. |
4.2 高级权衡句型
更高级的权衡句型用比喻 / 名词化——显得更文学性,适合 paper / 演讲。
| 句型 | 用法 |
|---|---|
| strike a balance between X and Y | We strike a balance between performance and cost |
| navigate the trade-off | The design navigates the trade-off via... |
| walk a tightrope between | The compiler walks a tightrope between speed and safety |
| the conflict / tension between X and Y | The tension between Si IGBTs and SiC MOSFETs... |
5. 研究语境关键词区分
下面 4 组词研究语境严格区分使用 —— 中国工程师常混用,显得 non-academic。
5.1 假设 / 推论类
中文里"假设"一个词覆盖很多——英文严格区分 hypothesis(待验证)/ assumption(已接受)/ proposition(论点)/ postulate(公理)。
| 词 | 含义 |
|---|---|
| hypothesis | 待验证的假说("We hypothesise that...") |
| assumption | 接受为真的前提条件("Under the assumption that...") |
| proposition | 提出的论点("The proposition is that...") |
| postulate | 公理 / 基本假设(更正式) |
5.2 结果 / 结论类
结果类 5 个词按"原始 → 综合"分层——observation 是最原始的现象,implication 是最综合的启示。每层都不能跳过。
| 词 | 含义 |
|---|---|
| observation | 测得的现象 |
| finding | 研究的具体发现 |
| result | 测试 / 实验的输出 |
| conclusion | 综合分析的总判断 |
| implication | 对其他领域的启示 |
例: "Our observation that X drops at high T led to the finding that Y is the cause. The results support our hypothesis. The conclusion is that Z. The implication is that all SiC drivers should consider..."
5.3 影响 / 关系类
关系类 5 个词按"弱 → 强"分层——association 最弱、causation 最强。误把 correlation 当 causation 是统计学最常见错误。
| 词 | 含义 |
|---|---|
| correlation | 统计相关(不蕴含因果) |
| causation | 因果关系(需要更强证据) |
| association | 关联(中性,弱于 correlation) |
| dependency | 依赖关系 |
| interaction | 相互作用 |
5.4 准确度类
准确度类 5 个词含义完全不同——accuracy 接近真值、precision 重复性、uncertainty 不确定度、tolerance 允许偏差、resolution 最小可分辨值。混用是测量学新手最常错误。
| 词 | 含义 |
|---|---|
| accuracy | 接近真值的程度 |
| precision | 重复性(不一定准确) |
| uncertainty | 测量不确定度 |
| tolerance | 允许偏差范围 |
| resolution | 最小可分辨值 |
注意:accuracy ≠ precision ——常见错误是把高 precision 的仪器当作高 accuracy。
6. 量化表达
技术写作必须量化——"显著提升"是 6 分卷,"10.3% improvement"是 8+ 分卷。下面 5 个量化句型让你精准表达。
| 句型 | 例 |
|---|---|
| a [number]× improvement | a 3× improvement in switching speed |
| a [percent] reduction in | a 28% reduction in turn-on energy |
| on the order of | losses on the order of 100 mW |
| within a factor of | accurate within a factor of 2 |
| less than [N]% of | leakage is less than 1% of nominal current |
6.1 数量级表达
数量级有 5 个标准短语——order of magnitude / negligible / substantial 等。这些是技术 paper 表达"大致大小"的核心词汇。
| 表达 | 含义 |
|---|---|
| an order of magnitude | 1 个数量级(10×) |
| two orders of magnitude | 2 个数量级(100×) |
| negligible | 可忽略不计 |
| substantial / non-trivial | 不可忽略 |
| marginal / modest | 微小但显著 |
7. 5 个 Tech Writing 反模式
技术写作失分集中在 5 个反模式——这 5 个让审稿人立刻识别 non-academic 作者,即使技术内容正确。
| 反模式 | 表现 | 修法 |
|---|---|---|
| 日常词代学术词 | "we made an experiment" | "we conducted an experiment" |
| 多层 of 嵌套 | "the system of the design of the inverter" | 用 noun chain |
| 缺量化 | "showed significant improvement" | "showed a 28% improvement" |
| 混用 hypothesis/assumption | 把假设当假说 | 严格区分研究语境词 |
| accuracy / precision 混用 | "high accuracy"(其实是 precision) | 严格区分 |
7.1 日常词代学术词的隐蔽危险
中国工程师用 "we made / did / got" 等日常词写 paper —— 字面对,但 register 错。审稿人立刻识别 non-academic。
修法:记住 30 个高频 verb-noun pair(本页 §1),写作时强制替换日常词。
7.2 缺量化的隐蔽危险
"显著提升 / 大幅减少" 是中文技术写作传统,但英文 paper 不接受——必须给具体数字。审稿人第一句就会问"how much?"。
修法:每个 claim 必有数字 —— 实验设计时就预设要测的量化指标。
核心要点
- 技术 verb-noun 30 个高频搭配:实验类(8) + 分析类(6) + 控制类(8) + 验证类(4) + 表征类(4)
- characterize ≠ parameterize ≠ specify —— 严格区分用法
- Noun chain 替代多层 of —— 把 "X of Y of Z" 改成 "Z's Y X" 更简洁
- 权衡句型 5 种:trade-off between / at the cost of / in exchange for / sacrifice X for Y / X comes at the expense of Y
- 研究语境词严格区分:hypothesis / assumption / observation / finding / conclusion / implication / correlation / causation / accuracy / precision
- 量化表达:必须给具体数字 —— 5 种量化句型 + 数量级表达
- 5 反模式戒除:日常词代学术词 / 多层 of 嵌套 / 缺量化 / 混用 hypothesis/assumption / accuracy/precision 混用